ABSTRACT

A FACET THEORY ANALYSIS OF "WHAT'S IN A NAME":
BLACK VERSUS NEGRO

BY

Martin George Brodwin

Statement of the Problem

 

The very existence of racism in our world signifies
the importance of this study. In the United States, prej—
udice of white peOple toward blacks needs to be thoroughly
investigated for change to be instrumented. The colleges
and universities are ideal institutions for imperative
changes in race relations to originate and be generated
through the society. Thus, a study of "attitude-behaviors"
of college students is a valuable undertaking.

The present study1 reviewed the literature behind
the Attitude Behavior Scale as instrumented by Jordan and
Hamersma (l969),and explored mmmaspecific areas of atti-
tude research that led to development of the scale. The
research is based upon Guttman's (1950, p. 51) definition

of attitude as a "delimited totality of behavior with

 

1Part of a larger cross-cultural study under the
direction of John E. Jordan, College of Education, Michigan
State University, East Lansing, Michigan, 48823.

Martin George Brodwin

respect to something," and Jordan's (1971a) conception of
"attitude-behavior" to connote this.
Two different attitude instruments were used in

this study; the Attitude Behavior Scale, and a twenty-item

 

Semantic Differential scale as developed by Osgood, Suci,
and Tannenbaum (1957). Use of the Semantic Differential
provided information on the relationships between race,

color labelling, word symbolism, and attitudes about the

attitude object in-situation.

Methodology

Three forms of the racial attitude scale, ABS-WB-G,
ABS-WN-G, and ABS-BW-G, were used in this study. These
scales evaluated white college students' "attitude-
behaviors" toward two racial labelling concepts, 'black'
and 'Negro.’ These two attitude object labels were used to
assess whether the words 'black' and 'Negro' elicited
different attitude-behaviors in white college students.
The groups tested were also given a Semantic Differential
measure of attitudes toward race, along with the Attitude
Behavior Scale.

Several hypotheses were analyzed using product
moment coefficients, multivariate analysis of variance
techniques, two-sample t-tests, and the Kaiser 22 for

simplex approximation.

Martin George Brodwin

Results

Some of the more important results were the

following:

1.

No differential relationships were found
between students taking the "ABS toward
blacks" and those taking the "ABS toward

Negroes,‘ on the six Levels of the

Attitude Behavior Scale.

 

The attitude data from the samples did
approximate a Guttman simplex structure.

Some significant correlations were found
between the variable of efficacy and the six
ABS Levels.

There were no significant correlations between
the 'evaluation,"potency,' and 'activity'
dimensions of the Semantic Differential and

the six Levels of the ABS.

A FACET THEORY ANALYSIS OF "WHAT'S IN A NAME":

BLACK VERSUS NEGRO

By

Martin George Brodwin

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling, Personnel Services
and Educational Psychology

College of Education

1973

© Copyright by
MARTIN GEORGE BRODWIN

1973

What shall I tell my children who are black

Of what it means to be a captive in this dark skin?

What shall I tell my dear one, fruit of my womb,

Of how beautiful they are when everywhere they turn

They are faced with abhorrence of everything that is black
The night is black and so is the boogyman.

Villains are black with black hearts.

A black cow gives no milk. A black hen lays no eggs.

Bad news comes bordered in black, mourning clothes black,
Storm clouds, black, black is evil

And evil is black and devils food is black . . .

 

 

What shall I tell my dear ones raised in a white world
A place where white has been made to represent

All that is good and pure and fine and decent,

Where clouds are white and dolls, and heaven

Surely is a white white place with angels

Robed in white, and cotton candy and ice cream

And milk and ruffled Sunday dresses

And dream houses and long sleek Cadillacs

And angel's food is white . . . all, all . . . white.

(Margaret Burroughg,cited in Banks & Joyce, 1971, p.35).

ii

Dedicated to my brother, Buddy

iii

PREFACE
This study is one in a series, jointly designed by
several investigators, as an example of the 'project'
approach to graduate research. A common use of instrumen-
tation and theoretical material, as well as technical and

analyses procedures, was both necessary and desirable.

iv

ACKNOWLEDGEMENTS

First of all, I would like to thank my advisor and
chairman Dr. John E. Jordan who has not only given me
guidance and encouragement in all aspects of this disser-
tation, but has helped give my life direction. With him,

I hope for a lifelong friendship and continuing professional
relationship.

Also, I would like to eXpress my deep appreciation
to my committee members: Dr. Thomas Gunnings, Dr. Maryellen
McSweeney, and Dr. Stephen Yelon for their time in reviewing
the data, making helpful suggestions, and giving me new
insight.

The roles of Dr. Keith Anderson and Dr. R. G. Rex
were paramount in the administration and gathering of my
data.

Also thanks to my friends and fellow students,
Bahman Dadgostar, Marie Hightower, Sue Kohen, and B. N.
Singh who aided in the completion of the data.

I would like to mention the names of those doctoral
students whose encouragements and friendships have helped
make my stay at Michigan State University an enjoyable

experience; John Aycock, Lynne Brody, Bahman Dadgostar,

Ken Gottlieb, James Hightower, Aubrey Radcliffe, B. N.
Singh, and Win Smith.

I would like to thank my friend and typist, Judy
Little, who helped me finish work on this thesis.

Last but by no means least, without the continuous
assistance and encouragement from my parents, Dr. and Mrs.
Allen L. Brodwin, and my brother Bennett, this dissertation

would not be possible.

vi

TABLE OF CONTENTS

PREFACE O O O O O O O O O O O O O 0
ACKNOWLEDGEMENTS . . . . . . . . . .
LIST OF TABLES O O O O O O O O O O O 0
LIST OF FIGURES . . . . . . . . . . . .
Chapter
I. INTRODUCTION . . . . . . . . . .
Problem . . . . . . . . . .
Need 0 O O O O O O 0 O O 0
Purpose . . . . . . . . . .
Definitions . . . . . . . . .
Hypotheses . . . . . . . . .
Organization of the Thesis . . . .

II.

III.

HISTORICAL ANTECEDENTS OF FACET THEORY AS
IT RELATES TO GUTTMAN-JORDAN METHODOLOGY .

Early Foundations of Attitude-

Behavior Research . . . . . .
Early Experimental Research, Theory and
Development: British Empiricism and
German Experimentalism . . . . .

The Development of Quantitative
Psychology . . . . . . . . .

Guttman Methodology . . . . .
Theoretical Developments in Attitude-
Behavior Research . . . . . .

The Conception of Attitude as
Attitude-Behavior . . . . . .
Summary . . . . . . . . . .
RACIAL ASPECTS OF PREJUDICE . . . . .

Introduction . . . . . . . . .
Race Labelling . . . . . . . .
The Relationship Between Language
and Racial Prejudice . . . .
Cognitive, Affective, and Behavioral
Aspects of Racial Attitudes . . .
Conclusion . . . . . . . . .

vii

Page

iv

ix

xii

15
l6
l7
l8

19

21

24

3O
51

76

83
93

95

95
96

110

114
115

Chapter

IV. RACE, COLOR-LABELLING, AND THE

SEMANTIC DIFFERENTIAL

Introduction . . .
Word Symbolism . .

The Semantic Differential Technique
Color Labelling and the Semantic

Differential . .

Conclusion

V. METHODOLOGY

VI. DESI

Review of Attitude Scales

Guttman's Theoretical Structure:

A Four-Level Theory .
Jordan's Six-Level Adaptation . .
Attitude-Behavior Scales (ABS) . .

The Simplex Approximation and the
Reliability and Validity

GN . .

Instrumentation . .

Procedure

Research Hypotheses

VII. ANALYSIS OF THE DATA .

VIII. SUMMARY, DISCUSSION, AND RECOMMENDATIONS

REFERENCES

APPENDICES

A.
8-1.
B-2.
B-3.
C.

Analysis Procedure .

Research Hypotheses

Summary of the Study
Discussion
Recommendations for Further Research
Summary

Glossary
Attitude
Attitude
Attitude
Semantic

of the ABS

Behavior Scale WB-G . .
Behavior Scale WN-G . .
Behavior Scale BW-G . .

Differential

viii

ABS
Scales

Page

118

118
119
121

130
147

150
153

156
162
173
188
199

204

204
204
210

221

221
224

237

238
241
256
265

269
286

287
292
313
334
355

Table

1.

lo.

11.
12.

13.

14.

LIST OF TABLES

Hierarchy of Coefficients (Amalgamated
Series). (A) Elementary School . . .

Hierarchy of Coefficients (Amalgamated
Series). (B) Preparatory School . . .

Correlations of Peterson . . . . . .

Correlation Matrices with Variables in
Hierarchical Order . . . . . . .

Correlation Matrix . . . . . . . .

Intercorrelations of Six Hypothetical Tests
Having One Common Factor, Illustrating
the Condition of Simple Proportionality
in a Correlation Matrix . . . . . .

Same Intercorrelations as in Table 6 with
Variables Rearranged so as to Show the
Proportionality More Clearly . . . .

Test Intercorrelations for a Hypothetical
Single-Common-Factor Structure . . .

Test Intercorrelations for a Hypothetical
Equally-Spaced, Perfect Simplex . . .

Test Intercorrelations for a Hypothetical,
Non-equally-Spaced, Perfect Simplex . .

Scalogram Diagram for Three Dichotomous Items

Facets on which Subuniverses Differ . .

Guttman Facet Profiles of Attitude
Subuniverse . . . . . . . . . .

Empirical Intercorrelation of Scores on the
Four Subuniverses . . . . . . . .

ix

Page

34

35

36

38

43

47

47

62

64

65
73

157

158

160

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

Page

Hypothetical Matrix of Level-By-Level
Correlations Illustrating the
Simplex Structure . . . . . . . . . . 161

Jordan Facets Used to Determine Joint
Struction of an Attitude Universe . . . . 163

Joint Level, Profile Composition, and Labels
for Six Types of Attitude Struction . . . . 164

Comparison of Guttman and Jordan Facet
Designations . . . . . . . . . . . 165

Joint Level, Profile Composition, and Labels
for Six Types of Attitude Struction . . . . 166

Permutations of Five Two-element Facets of
Table 16 . . . . . . . . . . . . . 168

'Combinations of Five Two-element Facets and

Basis of Elimination . . . . . . . . . 169

Five-Facet Six-Level System of Attitude
Verbalizations: Levels, Facet Profiles,
and Definitional Statements for Twelve
Combinations . . . . . . . . . . . 170

Analysis of Theoretical Correlations of
"Perfectly Ordered" Matrices with Equal
and Unequal Differences Between Correlations . 190

Analysis of Simplex Correlations of the ABS-MR
Test Development Data for the ED 200, Belize,
and SER Samples . . . . . . . . . . 192

Simplex Results for Research Groups on Six
Levels of the ABS-DF . . . . . . . . . 194

Analysis of Simplex Correlations of the ABS-
BW/WN-G for the Samples . . . . . . . . 195

Analysis of Simplex Correlations of the ABS:
BW/WN for the Research Groups . . . . . . 196

Correlation Matrices and g2 Values for
Original and Best Simplex Approximations
on the ABS-DU, All Categories, Initial Scale . 197

ABS-BW/WN/WB: Basic Variables List by IBM
Card and Column. Brodwin Study (U.S. = 133) . 222

Table Page

30. Multivariate Analysis of Variance (MANOVA)

Results for Racial Terms, 'Black' and

'Negro' O O O O O O O O O O O O O 226
31. Multivariate Analysis of Variance (MANOVA)

Results for the Three Administration Groups . 226

32. Correlation Matrices and 92 Values for
Original and Best Simplex Approximations . . 227

33. Correlations and Significance Levels of

Efficacy to the Six Levels of ABS: WB-WN . . 232
34. t-Score Values for Semantic Differential

Ratings of 'Negro Person' and 'Black Person' . 232
35. Correlations and Significance Levels of

Efficacy and Five Concepts of the
Semantic Differential on the Evaluation
Dimension . . . . . . . . . . . . 233

36. Correlation Coefficients and Significance
Levels Between the Six Levels of the ABS
and the 'Evaluation' Dimension of 'Black'
and 'Negro' of the Semantic Differential . . 235

37. Correlation Coefficients and Significance
Levels Between the Six Levels of the ABS
and the 'Potency' Dimension of 'Black'
and 'Negro' of the Semantic Differential . . 235

38. Correlation Coefficients and Significance
Levels Between the Six Levels of the ABS
and the 'Activity' Dimension of 'Black'
and 'Negro' of the Semantic Differential . . 235

xi

Figure

1.

LIST OF FIGURES

Page

Mapping Sentence for the Facet Analysis of

Joint and Lateral Dimensions of Attitude-

Behaviors Toward Participation in Specified

Activities . . . . . . . . . . . . 177
A Mapping Sentence for the Facet Analysis

of Attitudes Toward Mentally Retarded

Persons . . . . . . . . . . . . . 178
A Mapping Sentence of the Conjoint, Disjoint,

and Response Mode Struction Facets used to

Structure the Attitude Behavior Scale—

Mental Retardation . . . . . . . . . 179

Mapping Sentence for the Facet Analysis of
Joint and Lateral Structure of Blacks'
and Whites' Attitudes Toward Each Other . . 180

A Mapping Sentence for Strategies of
Theory Development . . . . . . . . . 184

xii

CHAPTER I

INTRODUCTION

What happens to a dream deferred?
Does it dry up

like a raisin in the sun?

Or fester like a sore--
‘And then run?

Does it stink like rotten meat?
Or crust and sugar over--

like a syrupy sweet?

Maybe it just sags
like a heavy load.

OR DOES IT EXPLODE
(Langston Hughes, 1966, p. 268).
Problem

Perhaps the most crucial issue facing Americans in
the United States today, concerns their feelings and beliefs
toward persons of differing backgrounds and cultures.
Relations between blacks and whites in this country are at
a critical crossroads; both races have become increasingly
outspoken in expressing their opinions toward members of
the other race. EXpression of these views has not always
been rational or representative, but nevertheless these
expressions stem from real feelings, emotions, hurts, and
misunderstandings.

The education and training going on in our schools
Ith only reflect the views of the surrounding society, but

1

can have an effect in changing them. Students now in
school will soon become a permanent part of the larger,
surrounding environment and culture. As one psychologist
sees it (Gunnings, 1971, p. 101), "our young students must
not become protectors of the system, but innovators who are
striving to make this a better world for all." If we as
educators, counselors, and researchers fail to make an
impact on our students, to at least make them aware of
racial injustice, we will have as a result contributed to
racism in American society. Students must become willing
"to take on the system--a system that has so obviously
created inferior inner city education, a system that has
dehumanized man, and a system that has oppressed minori-
ties" (Gunnings, 1971, p. 101).

America is at present a racist society. Racism and
racist conceptions pervade most of the institutions in
this country. Blauner (1970, pp. 115-116) pointed out two
distinctive characteristics of this racism. "First, that
(aside from age and sex) the division based upon color is
the single most important split within the society, the
body politic, and the national psyche." The second
characteristic contributing to racism, cited by Blauner,
involves the major institutional structures in America:
"that various processes and practices of exclusion and
subordination based upon color are built into the major

public institutions (labor market, education, politics, and

law enforcement) with the effect of maintaining special
privileges, power, and values for the benefit of the white
majority."
The black man in America has not been the only

group discriminated against. According to Allport (1954),
prejudice against all minority groups has had a similar
basis:

The rich take to opium and hashish, those who cannot

afford them become anti-Semites. Anti-Semitism

is the morphine of the small people . . . Since

they cannot attain the ectasy of love they seek the

ectasy of hatred . . . It matters little who it is

they hate. The Jew is just convenient. If there

were no Jews the anti-Semites would have to invent

them (p. 343) .
Out of the restrictions and policies of racism, grew a
demand for black identity, soon transformed into a quest
for a recognition of a broad1y~based black culture. Recog-
nition of this black culture by the white majority would
enhance understanding between the races. "For white
Americans, too, a recognition of the vitality and innova-
tion of all of Afro-America may help organization and
expression and give us all a fuller means of understanding
and prizing cultural diversity in our midst" (Szwed, 1970,
pp. 295-296). Szwed emphasized the importance for whites
to "grasp the dimensions of Afro-American culture and

history and restore complete identity to a people who have

been so long divided from their own past and from each

other" (p. 296).

The demand for whites to recognize and accept
African culture in black Americans has been noted by other
authors. However, there is an inherent confusion on the
part of whites in accepting the fact that blacks do have a
cultural heritage that they want to assert. For so many
years, white America has denied this cultural heritage to
blacks; it now becomes very difficult for whites to see
and accept it. They want to deny its existence, and often
feel that blacks should be assimilated into the white
dominant middle-class society. The view of blacks as
simply "white men in black skins" pervades much of white
and even black thinking (Fanon, 1967).

The struggle of the black man in America is
related to the world-wide changes taking place especially
in Africa. During the past twenty years, newly born
nations in Africa have been seeking and struggling for
independence. In his quest for 'black culture,‘ the
American black has been discovering a unity with these
nations. "In his own battle, the American Negro is able
to achieve a new sense of kinship and feeling of purpose--
a new, larger, black identity. The struggle of black men
has become symbolic of the struggle of all oppressed
groups to achieve dignity and respect in the face of bigotry
and discrimination" (Proshansky and Newton, 1968, p. 215).

A cultural revolution is a change of outlook. It takes

place inside the head. It is an overthrow of old
values, prejudice, beliefs, and opinions, and the

installation of the new. What happens outwardly
merely reflects this internal transformation.

This revolution among blacks in America shone
forth in the form of great self-pride, assurance,
fearlessness, determination, and a new aggressive-
ness (Smith, 1970, p. 38).

According to many social scientists, the words
'Negro' and 'black' have different connotations for both
the white community and the black community. The term
'Negro' is more associated with past racist notions of
inferred inferiority. 'Black' is more associated with
pride and power in being black, and in black heritage as
related to African ancestry.

Black Power is concerned with organizing the rage

of Black people and with putting new, hard questions
and demands to White America. As we do this, White
America's responses will be crucial to the questions
of violence and viability. Black Power must (1) deal
with the obviously growing alienation of black
peOple and their distrust of the institutions of
this society; (2) work to create new values and

to build a new sense of community and belonging;

and (3) work to establish legitimate new institu-
tions that make participants, not recipients, out

of a people traditionally excluded from the funda-
mentally racist processes of this country (Hamilton,
1969, p. 126).

The importance that terminology plays in the
changing identity of blacks in this country has been suc-
cinctly put by Killian and Grigg (1964): "at the present
time, integration as a solution to the race problems
demands that the Negro foreswear his identity as a Negro."

Recent literature (Jordan, 1971a; Hamersma, 1969)

points out that the black community has, as a whole,

accepted and preferred the term 'black' as opposed to

'Negro.‘ The white community, however, has been slower in
accepting this newer terminology, and many of the values
that accompany it.

The specific, experimental problem of this study
was to assess the differing attitudinal dispositions that
white college students associate with the word 'Negro' as
opposed to the word 'black' in an attitude scale.

Most attitude research has been of a theoretical
nature. Studies point out what they feel attitude is
composed of, but they rarely submit their theory to
statistical analysis. The Jordan-Guttman system is an
attempt to devise a definition of attitude that encompass
both the semantic, theoretical analysis and the statistical
structure underlying the definition. Jordan has eXpanded
Guttman's three-facet, four-Level theory, into a five-facet,
six-Level design, encompassing the old tripartite (Plato)
scheme of analysis by dividing attitude into cognitive

elements, affective elements, and conative elements.

13229
According to C. Eric Lincoln (1968) one result of
the present black power struggle has been an attack on the
word 'Negro.‘ Many black authorities consider this word
as originating from the time of slavery in this country and
insist on usage of the words 'black' or 'Afro—American.‘

The coordinator of the Afro-American History and Cultural

Center of the New York City Board of Education, Keith Baird
(cited in Lincoln, p. 132), stated that: "This is not a
minor semantic dispute. It engages the emotions and intel-
lect of a vast number of people, from Southern campuses to
the corner of 125th Street and Seventh Avenue in Harlem

(p. 132). Another scholar in black psychology, Fuller
(cited in Lincoln, p. 133), saw the issue in terms of a
generation gap: "those who are willing to accept the
'status quo' use the term 'Negro,‘ while those who seek
improvement use 'black' or 'Afro-American.‘ Such organi-
zations as the "National Association for the Advancement of
Colored People" and journals such as "Negro Digest" are
under much pressure to change their titles. M. R. Karenga
stated the following concerning black identification:
"Yesterday we thought we were Negroes. Today we know we
are Black men but we still have some Negro hang-ups." He
stressed the concept of culture giving people an identity,
a purpose, and a direction; culture "tells us (blacks) who
we are, what we must do, and how we can do it." Karenga
saw 'Negroes' as imitating white culture for so long that
they have begun to believe that it is their own (Lincoln,
p. 10-11). Black Power advocates use the word 'black' for
"black brothers and sisters who are emancipating themselves,"
while using the word 'Negro' contemptuously for Negroes
"who are still in Whitey's bag and who still think of

themselves as Negroes" (Bennett, 1967, p. 47).

Anything except that odious word, 'Negro' which has
such a ridiculous heritage of mockery. We were

not dropped into America; we were forcibly dragged
from Africa. So what else are we? Words are the
greatest avenue of communication and much more
reliable than signs. They are forceful symbols . . .
Let us all rally around Afro-American. I think
black has connotations of arrogance (J. Harden,

1968, p. 10).

J. Leo of the New York Times saw the recent effort
to find a more meaningful term beginning in the late 1950's
when many black nations in Africa sought independence. As
the myth of uncivilized Africa was corrected, the road was
open for identification with these newly emerging nations
(Lincoln, 1968).

Williams and Kirkland (1971) differentiated between
'black' and 'Negro' thusly:

To be Negro in America has certain meanings which
need to be clarified here. Negroes are made and
manufactured in the United States. The following
descriptions are representative: The Negro is
concerned with education for purposes of individual
achievement; he wishes to become integrated or
assimilated into the mainstream; he accepts white
standards; he is materialistic; he accepts gradual-
ism; he defines the problem as within the Negroes,
not in the system.

To be a black man in contemporary America has
certain connotations regarding self-definition:

The Black man sees education as a vehicle for social
change; he believes group goals are more important
than individual goals; nation building is vital

to his survival; he selects leaders based on compe-
tence, not status; he is pro-Black, not necessarily
anti-White; he sets norms and defines goals in

terms of self-determination; there is a sense of
urgency about his goals (p. 115).

R. Moore, in his book, The Name Negro Its Origin

 

and Evil Use, stated that the word 'Negro' is so 'saturated

with filth,‘ so 'polluted' with the white man's stereo-
types, that "there is nothing to be done but to get rid of
it" (Bennett, 1967, p. 54).

To many, the term 'Negro' means the continuation
of 'master-slave mentality' first perpetuated in the slave
period in America. "They maintain that a change in name
will short-circuit the stereotyped thinking patterns that
undergrid the system of racism in America" (Bennett, 1967,
p. 47).

According to D. Edwards, assistant managing editor
of the New York Amsterdam News, one of the largest black
newspapers in America, young blacks especially are against
continued use of the label 'Negro.' They often associate
it with the oppression and slavery into which blacks were
born, and the fact that this label was forced upon them
involuntarily. The word is thought of in connection with
"Uncle Tomism." 'Black' or 'African' is the preferred
usage (Bennett, 1967).

K. Baird felt that the identity crises blacks
undergo in Americavnmsheightened by use of the designation
'Negro.' It helps promote the continued depressed economic
and social status of blacks in America. As Baird stated:
"Positive and enhancing self-regard is a psychological
necessity of life, and the name borne by an individual or
group can be an effective vehicle and symbol of group or

individual self-regard" (Bennett, 1967, p. 52).

10

Language tends to prestructure both thinking and
acting. Baird was concerned with the effect certain words
have upon people: "A name can determine the nature of the
response given to it by virtue of the associations which
its use conjures up" (p. 52). Baird did not believe that
a change in vocabulary would solve all racial problems
but he feltijzcould make a significant difference in rela-
tions between blacks and whites. "The very act and fact of
changing the designation will cause the individual to be
redesignated to be reconsidered, not only in terms of his
past and his present but hopefully in terms of his
future . . . Designation has an important bearing on
destiny" (Bennett, 1967, p. 52).

The word came into use, Baird says, in connection

with the enslavement of the African in the New

World. The use of the word became connected with what
Earl Conrad has so well called the 'Negro-Concept,‘
that grotesque conception of the African which has
been shaped in the mind of the European and forced
with Procrustean cruelty on the person and person-
ality of the black American (Bennett, 1967, p. 52).

"Baird believed that the word, 'Afro-American,’
would soon supplant the word 'Negro.' He did not object
to the term 'black,‘ which, he said, lacks the historical
and cultural precision of the word 'Afro-American'"
(Bennett, 1967, p. 54). A second author, R. Moore, supported
Baird's philosophy. "Black," Moore stated, "is a loose

color designation which is not connected with land, history

and culture. While I (Moore) recognize it as a step

11

forward in getting rid of the term 'Negro,‘ I think it is
necessary to take the next step" (p. 54).

The word 'Negro' is of Portuguese and Spanish
origin, dating back to the time of the African Slave Trade,
where the term referred to those Africans captured and
transported as slaves to the New World. "This word, which
was not capitalized at first, fused not only humanity,
nationality and place of origin but also certain white
judgments about the inherent and irredeemable inferiority
of the persons so designated"(Bennett, p. 48). Literate
blacks of the time, preferred the terms 'African' and
'black' and resisted usage of the label 'Negro.' The
first institutions and organizations begun in America by
those of African origin carried African designations:
"The Free African Society," "The African Methodist Epis-
copal Church," "The African Baptist Church." The Free
African Society was founded in Philadelphia in 1787. Its
preamble began: "We, the Free Africans and their descen-
dants of the city of Philadelphia in the state of Penn-
sylvania or elsewhere . . ."(Bennett, p. 48).

Further resistance to the term 'Negro' can be
found in the 1868 Constitutional Convention of North
Carolina. James Walker Hood, one of fifteen black dele-
gates to that convention, expressed the belief that there
were no 'Negroes' present at the convention. Further, he

insisted "that the word 'Negro' had no significance as to

12

color, but could only be used in a reproachful or degrading
sense"(Bennett, 1967, pp. 48-50).

The present study was designed to test whether
racial terminology, in particular 'black' versus 'Negro,'
is related to the way in which white people perceive black
people. Research on the Semantic Differential (Williams,
1964, 1966) suggested that the term 'black' has been given
many negative connotations. It is unclear how much of the
negative connotations attributed to the word 'black' gen-
eralizes to the word 'Negro.' It would be useful to
discover more of the underlying reasons as to why the two
racial terms, 'black' and 'Negro,' elicit different
"attitude-behaviors," if indeed they do.

Since the data of this study was of a correlational
nature, inference as to the direction of causality would
not be feasible. One can formulate conclusions as to the
importance of color labelling as eliciting more positive
or more negative racial "attitude-behaviors" in college
students. If the terms 'black' and 'Negro' do elicit
different attitudes, the relationship between race
labelling and racial attitudes would appear to merit
further research.

Valentine (1971) stated that for a white counselor
to be successful with a black client, his client must be
bi—cultural. Williams and Kirkland (1971) felt that this

is less than ideal. White counselors work most successfully

13

with the 'Negro' part of the black client or the part that
desires to be or become white, but cannot work with the
part that is black or desires to become black.

Before attempting to change attitudes, it is
necessary to discover what attitudes exist and where they
occur. Jordan (1970) felt that the value of such research
lies in the possible contributions toward the understanding
and conceptualizing of the determinants and facets underly-
ing the attitude structure. According to Yuker (1965,

p. 15), "an attitude can be defined as one type of pre-
disposition toward behavior. Because this is so, by
finding out what a person's attitudes are, we can gain
information that will help us both to understand and to
predict a person's behavior." Mehrens and Lehmann (1968)
stated that attitudes are learned and because of this can
be changed if it becomes necessary. However, before this
process of change can occur, it is necessary to discover
the present status and existence of the attitudes.

The need for the study of racial attitudes in our
society is not only obvious, but imperative. Behavior of
blacks and whites toward each other in this country has
reached the point of callousness and the height of ridicu-
lousness. It is no longer excusable, in the opinion of
this author, to allow even one 'ounce' of racist behavior
to go unchallenged. Efforts must be made to combat

racism in every institution and at every level in this

14

society. We must abhor and combat prejudice and racism
within our society and within ourselves until it becomes
only a vestige of a once-dreamt nightmare.

Need for research in the area of attitudes is of
paramount importance at this time. Jordan's use of facet
analysis is an attempt to find a system that will facetize
attitude into its semantic and structural components.

Many studies have analyzed and defined attitude
and delved into the theory behind the concept. Since the
time of Plato (Allport, 1954), attitude has traditionally
been divided into three separate areas: thinking, feeling,
and acting (cognitive, affection, and conation). This
tripartite scheme has continued to the present time.

As theories of attitude were being formed, so was
the statistical branch of psychology. The factor analytic
methods owe their beginning to early Greek quantitative
methods and development of the scientific method. Research
and expansion of factor analysis was continued by Thurstone
(1935, 1947) and Spearman (1927, 1951). R. B. Cattell
(1952, 1964) was the first psychologist to apply factor
analytic methodology to analysis of personality theory.

Guttman (1944, 1953b, 1958) began the use of facet
analysis, a technique similar to the traditional factor
analysis of Thurstone and Spearman, but clearly distinct
from it. Guttman (1950a) had operationally defined

attitude as "a delimited totality of behavior with respect

15

to something" (p. 51). Jordan (1971a) used this definition
to link attitude and behavior, in his attempt at construc—
ting attitude scales that will measure "attitude-behaviors"
on six levels. Jordan's theory combines a modification of
Guttman's (1959) three-facet, four-Level system with theo-
retical notions of "attitude—behavior“ consistent with the
cognitive-affective-conative analysis dating back to early

Greek philosophy.

Purpose

The purpose of this dissertation is two-fold. The
first is to find out whether the two terms 'black' and
'Negro' are indeed associated with differing "attitude-
behaviors" in white college students. The ABS is a measure
of attitude-behavior along a cognitive-affective-conative
trichotomy. The present study is designed to test whether
the two terms, 'black' and 'Negro,' elicit different
"attitude-behaviors" as measured by the ABS/WN-B. It was
hypothesized that the "ABS toward Negroes" will elicit
more positive "attitude-behaviors" than the "ABS toward
blacks."

The Semantic Differential was used as a second
measure of attitudes. It was hypothesized that students
would respond more positively to the word 'Negro' on a

Semantic Differential than they would to the word 'black.‘

16

The purpose of this study was to find out whether
or not the word 'Negro' elicited more positive responses on
these two measures of 'attitude' than the word 'black.‘

The second, and perhaps more encompassing purpose,

was to review the theory behind the Attitude Behavior Scale

 

to give the present author a better understanding of a
system that may eventually help promote and uncover a
meaningful definition and measurement of attitude. Through
the definition of attitude by the method of facet analysis,
it is hoped that the realm of attitude can then be compre-
hended sufficiently well that attitude change can be pre—
dicted and "controlled." This "prediction and control" will
hopefully be used to help people understand why certain of
their behaviors are harmful not only to others, but to them-

selves as well.

Definitions

 

Guttman (1950a,p. 51) defined attitude as a
delimited totality of behavior with respect to something."
An attitude is not merely a "predisposition to behavior,"
as previous researchers have claimed (Allport, 1935) but
is also the behavior with respect to the attitude object.

Jordan (l971a,pp. 6-7) agreed with Guttman's (1950a)
definition of attitude, stating that it "is consonant with
a structural or facet theory approach to the study of

attitudes and behavior"; Jordan used the hyphenated term

17

"attitude-behavior" to denote his conception of attitude

as 'behavior.' Throughout this study, attitude will be
conceived of in this manner, as "attitude—behavior."

When 'attitude' or 'behavior' is mentioned, it should be
noted that the term is referring to the totality of
attitudes and behaviors, taken as a single, unitary concept,

unless otherwise stated.

Hypotheses

 

H-l: Whites taking the "ABS toward blacks" will
have significantly more negative attitudes
than whites taking the "ABS toward Negroes."

H-2: The attitude data from the sample will form
a Guttman simplex.

H-3: There is a positive relationship between a
high efficacy score and positive attitudes
on the ABS.

H-4: Subjects taking the Semantic Differential
will give the concept 'Negro person' a
significantly higher rating on the 'evalua-
tion' dimension of the Semantic Differential
than the concept 'black person.‘

H-5: Subjects taking the Semantic Differential
will give the concept 'black person' a sig-
nificantly higher rating on the 'potency'
dimension of the Semantic Differential than
the concept 'Negro person.‘

H—6: Subjects taking the Semantic Differential
will give the concept 'black person' a sig-
nificantly higher rating on the 'activity'
dimension of the Semantic Differential than
the concept 'Negro person.’

H-7: A higher efficacy score on the ABS will be
correlated with more positive scores on the
'evaluation' dimension of the Semantic
Differential for the concepts 'friend,’
'black person,‘ 'white person,‘ 'Negro
person,‘ and more negative scores for the
concept 'enemy.‘

18

H-8: There will be a positive relationship between
high scores on the 'evaluation' dimension of
the concepts 'black person' and 'Negro person'
of the Semantic Differential and positive
scores on the ABS.

H-9: There will be no significant correlations
between high scores on the 'potency' dimension
of the concepts 'black person' and 'Negro
person' of the Semantic Differential and posi-
tive attitudes on the ABS.

H-lO: There will be no significant correlations
between high scores on the 'activity'dimen-
sion of the concepts 'black person'and 'Negro
person' of the Semantic Differential and posi—
tive attitudes on the ABS.

Organization of the Thesis

 

Chapter I serves as an introduction to the thesis.
It includes a statement of the problem, the need, and the
purpose of the study. Also included within this first
chapter is a brief section of definitions and the hypotheses
tested. An extensive review of Guttman-Jordan facet design,
tracing it back to its origin, comprises Chapter II.
Chapters III and IV review the relevant literature in the.
areas of race relations between black and white persons,
and the Semantic Differential as it relates to race and
color, respectively. The general methodology used in studies
dealing with the Attitude Behavior Scale is discussed in
Chapter V. Chapter VI refers to the specific design and
methodology of the present study. The data and results are
analyzed in Chapter VII, while Chapter VIII includes summary
material, recommendations for futher research, and conclu-

sions of the study.

CHAPTER II

HISTORICAL ANTECEDENTS OF FACET THEORY AS

IT RELATES TO GUTTMAN-JORDAN METHODOLOGY

Biological scientists agree that all men belong to
one species, homo sapiens ("wise man"). If Arthur
Koestler and others are correct in suggesting that
man may prove to be an evolutionary mistake, man's
choice of a name for his species will seem to have
been not only immodest (see Pettigrew, 1964, p. 59)
but most inappropriate. This, however, is a judg-
ment that the future will have to render (E. E.
Baughman, 1971, p. 2).

 

In order to understand the methodological and psy-
chological aspects of attitude research as related to human
behavior, it is beneficial to trace the historical develop-
ment of scaling techniques and attitude research from its
origin. This analysis will concentrate on those develop-
ments important and crucial to methodology used by both
Louis Guttman in facet analysis (1959) and John E. Jordan
(1971a) in his research on attitude-behaviors toward various
attitude objects.

To understand behavior, scientists throughout his-
tory have employed various techniques to categorize human
actions, beliefs, and thoughts and a number of techniques
have been developed by which behavior can be observed and

described. These methods can be classified into three

19

20

general categories. Observation is concerned with viewing,
description, and analysis of individual and group behavior.
This approach can be found in early anthropological and
sociological research studies, and it is still employed
today. In the second technique, that of self-report, the
subject reports to the researcher what he is feeling,
thinking, or believing, and/or what he has actually done.
The verbalizations are then classified, and from these
categorizations, an attempt is made to analyze behavior.
This technique is still employed in areas of psychology
such as psychoanalysis, Gestalt psychology, and existen-
tialism. A third technique which can be employed is that
of measuring behavior through some external, methodological
tool.

Jordan's (1970, 1971a, 1971b) attitude-behavior
research and his development of the series of Attitude-

Behavior Scales is a combination of the second and third

 

techniques. The Attitude-Behavior Scale is a self-report
instrument, attempting to measure an individual's thoughts,
feelings, and overt behavior. It is an attempt to define
attitude-behavior,l first through categorization and des-
cription and second through quantification, measurement,

and prediction of behavior.

 

1"Two basic views permeate the literature on atti-
tude research; one defining attitude as a 'predisposition
to behavior', and the second emphasizing attitude as
'behavior.'" Jordan (l97la), however, believes that
attitudes and behaviors are not separate or disparate

21

The present analysis is concerned with the methodo-
logical and theoretical construction and analysis of con-
ceptions related to attitude-behavior. A current and pro-
lific research endeavor in the area of scaling and attitude
research has been that of Guttman's facetized design and
scaling methodology as well as Jordan's attitude-behavior
scale analysis. This chapter traces the historical and
theoretical development of statistical innovations and
psychological interpretations and explanations of attitude-
behavior.

Early Foundations of Attitude-
Behavior Research

 

 

The early Greek philosophers laid the basic founda-
tion for what was eventually to become the 'scientific
method.‘ The cosmologists were the earliest known group
of thinkers who attempted to explain the environment that
surrounded them; their explanations revolved around under-
standing through animism, myth, and magic. According to
Hutten (1962), "Science begins with Thales . . . he exemp-
lifies the scientist who makes a bold, unifying hypothesis
but whose imagination is kept in bounds by a respect for

reality" (p. 57). The beginnings of the movement away

 

entities, but are varying points along the same variable;
hence, he used the hyphenated term, attitude-behavior, to
connote a synthesizing of what is usually considered two
separate and distinct entities. The new usage was partly
derived from Guttman's (1950a,p. 51) definition of atti-
tude, as a "delimited totality of behavior with respect to
something" (Jordan, 1971a, p. 7).

22

from mythical explanations of the world toward more scien-
tific, naturalistic observation can be attributed to the
time of this early Greek philoSOpher-scientist, Thales.

The most extensive reaction to the magic and mythical
interpretations of the cosmologists was that of the Sophists.
The beginnings of empirical research date back to the Greek
Sophist movement during the Golden Age of Greek civiliza-
tion: 461-431 B.C. They desired to study the individual
and his culture in a practical, concrete manner; they used
the empirical-deductive method as their main technique of
investigation. With this theoretical development, eXplana-
tions of human behavior became more observational and
empirical, replacing many of the past mythical notions and
conceptions.

According to Zeller (1881), however, the scientific
method cannot be credited specifically to Sophist philosophy;
their importance lies in breaking with the cosmological
explanation of the universe, which was based on myth and
religious foundations. Although the Sophists questioned
mythical conceptions, they did not systematically use
objective and empirical analysis. The Sophists prepared

the ground for the zeitgeist of scientific and philosophical

 

thought, that would permit the development of methodological
techniques and theoretical views through which the evolution

of the scientific method could be developed.

23

. . . for its purpose the Sophist movement was not
to guarantee objective knowledge, but only subjec-
tive readiness of thought and practical versatility.
This form of culture is tied to no scientific system
and principle, its distinctive character appears far
more in the ease with which it takes from the most
various theories whatever may be useful for its
temporary purpose; and for this reason it propa-
gates itself not in separate and exclusive schools,
but in a freer manner, by mental infection of
different kinds (Zeller, 1881, p. 514).

Stace (1967) interpreted this period of Greek
development as a time when the Sophists undermined and des-

troyed the beliefs in the classical interpretation of the

 

universe according to the cosmologists, and prepared the

way for such new interpretations of the weltanschauung of

 

Socrates, Aristotle, Archimedes, and others. It was
Aristotle who attempted to catalogue knowledge in a system-
atic form. Archimedes "anticipated the modern modes of
scientific thinking in his way of dealing with general
principles of nature" (Boring, 1950, p. 6). Thus, the later
Greeks, successors of earlier Greek contributors, began the
movement toward measurement and evaluation of the surround-
ing environment through quantitative methods.

Modern quantitative methodology dates back to the
four basic stages in the scientific method, developed by
the Greeks: (a) naturalistic observation, (b) classifica-
tion and analysis of natural phenomena into meaningful
descriptive categories, (c) formulation of hypotheses of
cause and effect based on such analyses, and (d) the value

of quantitative methodology. The Greeks extended their

24

studies into what can be considered the first applications
of experimental hypothesis testing and critical observa-
tion. "It is to the great credit of these ancient peOple
that they were able to develop what in essence amounts to
a sophisticated scientific methodology more than 2,000
years ago" (Chaplin and Krawiec, 1968, p. 17). The methods
of critical observation were prevalent in the West in the
Thirteenth Century with the rediscovery of Aristotle
(Crombie, 1952). The early British and German scientists
in the Eighteenth Century greatly enhanced and improved on
these first attempts at scientific methodology and experi-
mentation.

Early_Experimental Research! Theory and

Development: British Empiricism
and German Experimentalism

 

 

 

In search for the first experimentation that cul-
minated with the eventual development of a systematic and
methodological study of attitudes, the importance of early
Greek thinking has been noted. The next prominant develop-
ment in the trend that would eventuate in modern scaling
techniques involved British Empiricism. This school was
heavily steeped in the empirical tradition; the principle
of association, first seen in Aristotelian notions, was
extended by the empiricists. The general law of association,
first stated by early Greek philoSOphers, was taken up by
Hobbes and Locke in the Seventeenth Century and rediscovered

years later by Hartley:

25

The general law of association is that if sensa-

tions have often been experienced together, the

corresponding ideas will tend to occur together;

if A has been associated with B, C, and D in

sensory experience, the sensory experience A,

occurring alone, will tend to arouse the ideas of

B, C, and D, which accompanied it (Heidbreder,

1933, p. 54).
This law laid the foundation for what was later formulated
by Guttman (1959) as the 'contiguity hypothesis.‘ AS will
be seen, Guttman's hypothesis consisted essentially of a
modification and reformulation of this associationistic
principle applied to different realms of science, in par-
ticular that of intellectual ability and attitude research.

In Nineteenth Century Germany, new developments and

theories were introduced into the scientific stream of
thought which would permit the kind of research developed
by Guttman (1959) and Jordan (l971a). The fountainhead
of the mathematical and statistical developments in German
psychology can be traced to Kantian philosophy. Kant's
(Peters, 1962) first crucial contribution to the German
tradition of psychology, was to explicate the methods and
techniques of science; his second contribution "was his
contention that science is characterized by mathematical
as well as by empirical description. His celebrated fusion
of the empirical standpoint of Hume with the rationalist
standpoint of Wolff involved the aphorism that an empirical

inquiry is as scientific as it contains mathematics"

(p. 533). According to Brett, (Peters, 1962) Kant molded

26

and directed psychology into the area of measurement which
stimulated and enhanced the development of statistical and
mathematical models.

Two basic trends in psychology emerged at this time;
statistical-empirical on the one hand, and introspective-
action on the other. The beginning of the statistical and
mathematical approach to attitude measurement began with
these early German experimentalist and British empiricist
contributions. The concepts of Kant served as the corner-
stone to the whole school of experimental psychology in that
these researchers began to employ measurement and methodolo-
gical techniques. In close parallel to Kant was the eXperi—
mentalist, Herbart, who not only tried to formulate mathe-
matically precise laws of consciousness, but evolved tech-
niques whereby psychology, as a science, could employ the
mathematical model (Peters, 1962).

The experimental tradition continued with the work
of Weber (Chaplin and Krawiec, 1968), who served as a link
between the first early attempts at experimental investi-
gation and the beginning of experimental work in the field
of psychology that lead to eventual application of these
early mathematical techniques to the era of scaling and
attitude research. The first major application of this
new scientific methodology in the area of scaling and atti-
tude research, was undertaken by Fechner. According to

Chaplin and Krawiec (1968, p. 40), his "methods have stood

27

the test of time to become fundamental procedures in psycho-
physical measurements, mental testing, and attitude
scaling . . ."

Unlike Kant, Fechner was not a strict methodologist;
he had a humanistic philOSOphical bent along with his mathe-
matical and scientific interest. His significance lies in
the fact that he applied rigorous methods to practical,
everyday functioning. This is the crux of attitude research
today - to be able to make statements about behavior and
behavioral change, but at the same time to be able to make
these statements with some degree of validity and reliabil-
ity to relevant public concerns.

According to Brett (Peters, 1962, p. 534) "the main
function of measurement in science is surely to facilitate
the testing of hypotheses by expressing them more exactly.
Quantitative techniques enable scientists to answer
precisely questions unearthed by cruder qualitative methods."
This was what Fechner (Peters, 1962) attempted to do, and
also what Guttman and Jordan are attempting a century and a
half later. Both these attempts have employed rigorous
scientific methods to understand human thought and behavior.
As Brett (Peters, 1962, p. 534) has pointed out:

there is little point in going round measuring unless
the object of devising such measuring techniques is
the testing of interesting hypotheses. Measurement
by itself does not produce scientific hypotheses any
more than do laboratories or grants for research . . .
The advance of science depends upon the development

of imaginative assumptions as well as upon exact
techniques for testing them.

28

The next crucial step in the development of the
psychological measurement of attitudes is found in the con-
tribution of Wundt (Boring, 1950), considered the founder
of psychology as a formal discipline. Wundt laid the
foundation for modern experimental psychology. His use of
experimental observation and analysis for understanding
mental phenomena was the culmination of the trend begun
by Weber and Fechner to utilize physiological and physical
methods in psychological investigation.

Wundt provided an impetus for a new type of psy-
chology: that of the 'new' psychology of content, other-
wise understood as structural psychology. "He provided for
the new psychology its structure and form, its self-conscious-
ness, its name, its first formal laboratory, its first
experimental journal, as well as the systematic pattern with
respect to which the experiments could be formulated and
given their significance" (Boring, 1950, p. 334).

The 'new' psychology of content can be considered
introspective, sensationistic, elementistic, and associa-
tionistic. Consciousness was its subject matter; therefore,
it was intrOSpective. The nature of consciousness was
revealed through sensation - thus the 'new' psychology was
of a mental chemistry. Sensations, images, and feelings
were thought to be the elements making up the compounds
of mental thought. Lastly, because association was the very

principle of compounding, the 'new' psychology was

29

associationistic. The British Empiricists had previously
shown how perceptions and meaning can be obtained through
an association of various parts. This was the early law of
association postulated by Hobbes and Locke; its basic
conception was carried through psychological thought and is
seen in Guttman's (1959) 'law of contiguity' (Chaplin and
Krawiec, 1968).

Titchener (Peters, 1962), a strict student of
Wundt's, brought this 'new' psychology to America in the
form of Structuralism. The idea and practice of rigorous
analysis was carried through in psychology, thus making
psychology more scientific and respected by the physical
sciences. Titchener tightened the theory and experimenta-
tion of Wundt, while stressing the concept of 'structor'
by adding a new element, affective states, to the existing
states of sensations and images. These three elements,
Titchener utilized in classifying what he perceived as
conscious thought. All varied and complex mental processes
were derived from these three elements and their attributes
of quality, intensity, and duration. Titchener's structural-
ism gave psychology a more strict, rigorous scientific
flavor, that aided in the development of more exacting
measurement techniques, necessary for the birth and matu-

ration of attitude research.

30

The Development of Quantitative Psychology

 

John Graunt can be considered the first statistician
and the founder of statistics. In 1662, he published a
demographic analysis of plague deaths in London, perhaps the
first attempt to interpret biological and social phenomena

from quantitative data. His book, Natural and Political

 

Observations Made Upon the Bills of Mortality, was well

 

accepted as an important study of vital statistics; was
published and revised several times by Graunt; and was
published and enlarged by Sir William Petty after Graunt's
death.

Quetelet, a Belgian astronomer, made one of the
first successful attempts to apply statistical methods to
data involving human biological and social functioning. In
discussing human variability along certain variables,
Quetelet described the significance of the normal curve
distribution (Boring, 1950; Chaplin and Krawiec, 1968).

Galton used the work of Quetelet in developing
various methods of statistical analysis, such as the median,
standard score, and correlation technique. "The modern
techniques for establishing the validity and reliability of
tests, as well as the various factor analytic methods, are
direct outgrowths of Galton's discovery" (Chaplin and
Krawiec, 1968, p. 516). Galton was the first to formulate
the statistical tool of 'coefficient of correlation.‘

With Dickson, Galton published a paper in 1866 describing

31

what they called the 'index of co-relation.‘ Soon there-
after, it was renamed 'Galton's function,‘ and eventually,
in 1892, was changed to its present name, 'coefficient of
correlation'--symbolized by 'r'—-(Boring, 1950, p. 479).

The mathematical foundation of correlation can be
credited to the mathematician, Karl Pearson. In 1896,
Pearson used correlation in solving problems posed by mathe-

matical research in psychology and biology. Biometrika

 

was founded by Pearson, Galton, and Weldon in 1901. Both
Pearson and Galton, working together much of the time,
established statistical methods as a fundamental technique
for the investigation of psychological problems (Boring,
1950).

Just after these innovations, Spearman developed a
two-factor theory of human intellectual functioning
based on these newly discovered methods of correlation.
The two-factor theory interpreted intelligence as containing
a basic overall component, the general (G) factor, common
for all intellectual skills and activities, and several

specific (31, s .) factors, which varied for different

2 . .
skills and tasks of intelligence. In 1912, along with
Hart, Spearman devised a correlation matrix hierarchy for
these factors (Boring, 1950). This factor analytic work,
plus the advancements made especially by Thurstone have

culminated in Guttman's faceted definition of intellectual

ability and attitudes (Jordan, 1972b).

32

"We may note the line of descent for factor analysis,
noting only the prominant ancestors: Laplace - Quetelet -
Galton - Pearson - Spearman - Thomson - Garnett - Burt -
Thurstone" (Boring, 1950, p. 481). Although not an adher-
ent to factor analytic methods, Guttman (1948) used much of
the research advanced by the factor analysts. Their dis-
coveries in factor analysis and correlation matrix hier-
archies "laid the groundwork" for Guttman's facet theory and
structural analysis procedures such as the simplex.

Factor analysis is a method for analyzing a set of
intercorrelated performances into as many independently
variable factors as justify the labor of computation.
Each factor is defined by the degree to which it
participates in each of the various original perfor—
mances. You get the most important factor analyzed
first, and presently you stop with some residuals
that are too small to merit consideration. This
technic is used mostly with mental tests and is not
appropriate when the problem-situation can be
separated in advance into various parameters which
are subject to independent eXperimental control and
variation (Boring, 1950, p. 481).
At this last stage of separating the problem-situation in
advance into various parameters which are subject to inde-
pendent experimental control and variation, Guttman departed
from factor analysis by using facet methodology for ex-
ploring underlying 'dimensions.‘

The concept of hierarchical correlation as applied
to intelligence was influential in psychological research
around the turn of the century. Burt (1909) published one

of the first studies utilizing a hierarchy of correlation

coefficients. He stressed the belief that well measured

33

and controlled, applied data could provide just as sound or
even superior hierarchical coefficients of correlation than
theoretical data did. Spearman (1927) believed that a good
fit could be obtained only from theoretical coefficients;
Burt, however, attempted to prove that it was possible to
demonstrate a sound hierarchy of coefficients through
applied, well-controlled measurable data (see Tables 1 and
2). In a discussion of hierarchies, Burt stated the follow—
ing concerning Spearman: "Dr. Spearman and Prof. Krueger
imply that satisfactory hierarchies are exhibited only by
the 'pure' or theoretical coefficients, but it appears that
those based on amalgamated measurements are better than
those based on theoretical 'correlation', if the experimental
are carefully controlled" (Burt, 1909, p. 163). The corre-
lations do not fit the proposed scheme with perfect pre-
cision and cannot be expected to because like all empirical
observations they are subject to error.

The concept of hierarchical correlation was used
by several psychological researchers: (Peterson, 1908;
Stockton, 1921; Herring, 1921). Peterson employed (Table 3)
the statistical concept of hierarchical correlations in his
five-level table of intellectual ability (Spearman and
Jones, 1951).

During the early 1900's, statistical and psycho-
logical research and theory began noting the importance of

correlation coefficients and hierarchical analysis; however,

3&1

 

 

TABLE l.--Hierarchy of Coefficients (Amalgamated Series).a (A) Elementary School.
m
2
a. 3.
c\u w m 'u-« o c o
52 43 .5 s: .5 3.5 s .0 m 2 a z
u m n u s m H u u a. u c o o m
u Q a u o.u m o u a. u s c 5 g -H
on. H o E: a) 0.60 a: --4 o «a O (D
o d < m H H D m a e z m A 9 2 3
Dotting Observed coefficient - 77 67 60 69 57 57 50 52 48 38 20 16
Apparatus Theoretical value -- 80 73 72 72 67 63 49 45 33 28 27 05
Deviation -- 03 06 12 03 10 06 01 07 15 10 07 ll
P.e. of coefficient -- 05 07 O8 O6 O8 08 09 09 09 11 12 12
Alphabet Observed coefficient 77 -- 74 61 66 59 54 29 52 16 62 31 07
Theoretical value 80 -- 69 69 69 65 6O 46 43 32 26 25 05
Deviation 03 -- 05 08 O3 06 06 17 O9 16 36 06 02
P.e. of coefficient 05 -- 06 08 07 08 09 11 09 12 07. 10 12
Sorting Observed coefficient 67 74 -- 52 72 45 61 34 52 14 22 19 23
Theoretical value 73 69 -- 62 61 59 54 42 39 28 24 23 04
Deviation 06 05 -- 10 ll l4 13 08 13 14 02 04 19
P.e. of coefficient 07 O6 -- O9 06 10 08 11 09 12 ll 10 19
Imputed Observed coefficient 60 61 52 -— 44 76 47 67 4O 29 13 57 -13
Intelli- Theoretical value 72 69 62 -- 69 58 53 41 39 28 23 23 04
gence Deviation 12 08 10 -- 16 18 06 26 01 01 10 34 17
P.e. of coefficient 08 08 O9 -- 10 05 10 07 10 08 12 08 12
Dealing Observed coefficient 69 66 72 44 -- 76 47 67 40 29 13 S7 -13
Theoretical value 72 69 61 60 -- 58 53 41 39 28 23 23 04
Deviation O3 02 11 16 —- 07 12 01 05 19 00 04 03
P.e. of coefficient 06 07 O6 O9 -- 10 07 11 12 10 ll 12 12
Spot Observed coefficient 57 59 45 76 51 -- 41 41 47 25 63 26 11
Pattern Theoretical value 67 65 59 58 58 -- 48 37 35 35 26 21 04
Deviation 10 06 14 16 07 -- 07 04 12 01 18 05 07
P.e. of coefficient 08 08 09 05 09 -- 10 10 10 ll 12 ll 12
Tapping Observed coefficient 57 53 61 47 65 41 -- 41 47 08 26 -05 22
Theoretical value 63 6O 54 53 53 48 -- 36 34 25 20 20 04
Deviation O6 O6 O7 08 12 O7 -- 05 13 18 06 25 18
P.e. of coefficient 08 O9 08 10 08 10 -- 10 10 12 11 12 12
Mirror Observed coefficient 50 29 34 67 40 45 45 -- 34 16 08 05 -05
Theoretical value 49 46 42 41 41 37 36 -- 25 19 15 15 03
Deviation 01 17 08 26 01 04 05 -- 09 03 07 10 O8
P.e. of coefficient 09 11 11 17 10 10 10 -- 10 12 12 12 12
Sound Observed coefficient 52 52 52 4O 34 47 47 34 -- -O7 -01 01 -13
Theoretical value 45 43 39 39 39 35 34 25 -- 17 14 14 02
Deviation 07 09 13 01 05 12 13 09 -- 24 15 13 15
P.e. of coefficient 09 09 09 10 17 10 10 12 -- 12 13 12 12
Lines Observed coefficient 48 16 14 29 47 25 08 16 -07 -- 26 06 19
Theoretical value 33 32 28 28 28 26 26 25 17 -- 10 10 02
Deviation 15 16 14 01 19 01 17 03 24 -- 16 04 17
P.e. of coefficient 09 12 12 08 10 11 12 12 12 -- 11 12 12
Touch Observed coefficient 38 62 22 13 23 O3 26 08 -Ol 26 -- 16 29
Theoretical value 28 26 24 23 23 21 20 15 14 10 -- 08 01
Deviation 10 36 02 10 00 18 O6 07 15 16 -- 08 28
P.e. of coefficient 11 O7 12 12 12 12 11 12 12 11 -- 12 11
Memory Observed coefficient 20 31 19 57 19 26 -05 05 01 O6 16 -- 05
Theoretical value 27 25 23 23 23 21 20 15 12 10 18 -- 01
Deviation 07 06 04 34 04 05 25 10 13 04 08 -- 04
P.e. of coefficient 12 10 11 10 12 11 12 12 12 12 12 -- 12
Weight Observed coefficient 16 07 23 -13 01 11 22 -05 -13 19 29 05 --
Theoretical value 05 05 04 O4 04 O4 04 O4 03 03 01 01 --
Deviation 11 02 19 17 03 17 18 08 15 17 28 04 --
P.e. of coefficient 12 12 12 12 12 12 11 12 12 12 11 12 --
Average deviation = .100.
Average p.e. = .101.

 

aFrom Burt (1909, p. 161).

TABLE 2.~~Hierarchy of Coefficients

35

(Amalgamated Series).a

(B) Preparatory School.

 

 

8
a, 5
cxs E ro-S c cu m m
.52 a 3: a a :5 .2 . u. z c .5
sea a}: t: 2 3:: s: ‘d 5 2 .9: a z:
o a a E c -a o a m m o o -a w o m
0:: 4 HH 5: 2 m0. 6* m U) s-l 3 B Q
Dotting Observed coefficient ~~ 84 84 71 69 62 48 73 48 25 O7 03 ~03
Apparatus Theoretical value -~ 85 80 76 70 66 66 60 48 39 1 ~07 ~13
Deviation ~~ 01 04 05 01 04 16 13 00 14 07 10 10
P.e. of coefficient ~~ 06 06 10 12 12 16 10 16 19 20 2O 20
Alphabet Observed coefficient 84 -~ 80 48 84 67 57 76 34 22 ~14 ~28 45
Theoretical value 85 ~~ 78 74 68 64 64 58 51 37 14 ~07 ~12
Deviation 01 ~~ 02 26 16 03 07 18 17 15 28 21 57
P.e. of coefficient 06 -~ O7 16 06 15 14 O9 l8 19 20' 19 16
Imputed Observed coefficient 84 80 ~- 54 78 75 43 56 37 17 ~19 ~06 29
Intelli- Theoretical value 80 78 ~— 70 64 60 6O 55 44 35 13 ~06 ~12
gence Deviation O4 02 ~~ 16 14 15 17 Ol 07 18 32 00 41
P.e. of coefficient 06 07 ~~ 14 O8 O9 16 14 17 2O 19 20 18
Mirror Observed coefficient 71 48 54 ~~ 43 38 75 34 57 54 44 31 ~44
Theoretical value 76 74 70 -~ 61 58 57 52 42 34 12 ~06 ~11
Deviation 05 26 16 ~- 18 20 18 18 15 20 32 37 33
P.e. of coefficient 10 l6 l4 ~~ 16 17 09 18 14 14 16 18 16
Memory Observed coefficient 69 84 78 43 ~- 74 54 64 17 28 ~05 ~35 03
Theoretical value 70 68 64 61 ~- 53 53 48 39 31 11 ~06 ~10
Deviation 01 16 l4 18 -~ 21 01 16 22 O3 16 29 13
P.e. of coefficient 11 16 18 16 ~~ O9 14 11 20 19 20 18 20
Spot Observed coefficient 62 67 65 38 74 ~~ 38 51 25 34 07 ~44 19
Pattern Theoretical value 66 64 60 58 53 ~~ 50 45 36 29 11 ~05 ~10
Deviation O4 O3 15 20 21 ~~ 12 08 ll 05 O4 39 29
P.e. of coefficient 12 15 O9 17 09 ~- 17 15 19 18 20 16 19
Tapping Observed coefficient 48 57 43 75 54 38 ~~ 48 28 44 34 07 ~31
Theoretical value 66 64 60 57 53 50 ~- 45 36 29 11 ~05 ~09
Deviation 16 O7 17 18 01 12 ~~ 03 08 15 23 12 22
P.e. of coefficient 16 14 16 O9 14 17 ~~ 16 19 17 18 20 19
Sorting Observed coefficient 73 76 56 34 64 51 48 -- 38 00 ~22 ~14 02
Theoretical value 60 58 55 52 48 45 45 ~~ 33 27 10 ~05 ~09
Deviation 13 18 01 18 16 O6 03 -~ 05 27 32 09 ll
P.e. of coefficient 09 08 14 18 11 15 16 ~~ 17 20 19 16 20
Sound Observed coefficient 48 34 37 57 17 25 28 38 -- 07 34 17 ~17
Theoretical value 48 51 44 42 39 36 36 33 ~~ 21 08 ~04 ~07
Deviation 00 17 O7 15 22 11 08 05 ~~ 14 26 21 10
P.e. of coefficient 16 18 17 14 20 19 19 17 ~~ 20 19 20 20
Lines Observed coefficient 25 22 17 54 28 34 44 00 07 ~~ 35 19 ~13
Theoretical value 39 37 35 34 31 29 29 27 21 ~~ 06 ~03 ~06
Deviation 14 15 18 20 03 05 15 27 14 ~~ 29 22 O7
P.e. of coefficient 19 19 20 l4 19 18 17 20 20 ~~ 18 19 20
Weight Observed coefficient 07 ~14 ~10 44 ~05 O7 34 ~22 34 35 ~~ 38 ~35
Theoretical value 14 14 13 12 ll 11 ll 10 08 06 ~~ ~01 ~02
Deviation 07 28 32 32 16 04 23 32 26 29 ~~ 39 33
P.e. of coefficient 20 20 19 16 20 20 18 19 19 18 ~~ 17 18
Touch Observed coefficient 03 ~28 ~06 31 ~35 ~44 07 ~14 17 19 38 ~~ ~48
Theoretical value ~07 ~07 ~06 ~06 ~06 ~05 ~05 ~05 ~04 ~03 ~01 -- 01
Deviation 10 21 00 37 29 39 12 09 21 22 39 ~~ 49
P.e. of coefficient 20 19 20 18 18 16 20 16 20 19 17 ~- 15
Dealing Observed coefficient ~03 45 29 ~44 03 19 ~31 02 ~17 ~13 ~35 ~48 ~~
Theoretical value ~13 ~12 ~12 ~11 ~10 ~10 ~09 ~09 ~07 ~06 02 01 ~-
Deviation 10 57 41 33 13 29 22 11 10 07 33 49 ~-
P.e. of coefficient 20 16 18 16 20 19 19 20 20 20 18 15 ~-

Average deviation = .165.
Average p.e. = .162.

 

aFrom Burt (1909, p. 162).

36

 

 

TABLE 3.~~ Correlations of Petersona

Nature of Tests 1 2 3 4 5
Reasoning ~~ .95 .83 .40 .45
Generalization .95 ~~ .86 .40 .28
Abstract thought .83 .86 ~~ .64 .48
Memory .40 .40 .64 ~- .31
Accuracy .45 .28 .48 .31 ~-

 

aFrom Spearman and Jones (1951, p. 64).

these techniques would not gain prominance in the field of
attitude research until the early 1950's. Until statistical
theory and measuring techniques could be developed, and
practical use made of sampling methods, ordering of data,
and sampling error, further development of the hierarchy or
'order' concept could not be undertaken.

"Spearman differed from other intelligence testers
in trying to generalize the methods of factor analysis,
which had a practical origin, to the field of genera1 psy-
chological theory" (Peters, 1962, p. 734). Because of
Spearman's attempts to apply his research to fields other
than mental abilities, he has become extremely crucial and
relevant in the area of attitude research. "The statistical
approach of the Spearman School has also been applied to the
measurement of personality traits, attitudes, and values"
(Peters, p. 736). McDonnel (1927, cited in Peters) employed
these methods in studying bodily dimensions; Gates (1927,

cited in Peters) used them in a study of various physical

37

traits. In evaluating and diagnosing personality traits
and clusters, Eysenck relied on the statistical innovations
of Spearman (Peters, 1962).

These correlational techniques were also important
in Adorno, Frenkel-Brunswik, Levinson, and Sandford's work

on The Authoritarian Personality, and in Murray's Explora-

 

tions in Personality (Peters, 1962). Guttman (1958)

 

credited Spearman as being one of the forerunners to facet
theory analysis. Spearman postulated that "the mentally
presenting of any two characters tends to evoke a knowing
of the correlative characters" (Peters, p. 735). Basically,
this was a restatement of Locke's principles; the basic
notion of contiguity will later be seen in Guttman's work.
Although Spearman's laws were perhaps of questionable value,
it was his contributions to the conception of concomitant
variation between factors that became significant to the
field of psychology. "The discovery of concomitant varia-
tions is the beginning of the establishment of scientific
laws or relations of functional dependence between variables"
(Peters, p. 736). John Stuart Mill dealt with the issue of
concomitant variation between factors on a philosophical
basis before Spearman.

According to Cattell (1952), the birth of multi-

factor analysis began with Spearman, who was the first to

 

1Personal communication with Dr. Maryellen McSweeney,
College of Education, Michigan State University, Oct. 5, 1972.

38

develop theorems in factor analysis in his work with the
concept of intelligence as a single-factor theory. Spear-
man's hierarchical arrangement involved the entire matrix;
all adjoining columns in the matrix were proportional.

This is illustrated in Table 4, "Correlation Matrices with

Variables in Hierarchical Order."

TABLE 4.--Correlation Matrices with Variables in Hierarchical

 

 

Ordera

V6 V1 V4 V5 V3 V2 V3 V7
V6
V1 .90
V2 .82 .75
V3 .73 .61 .58
V4 .51 .49. .43 .36
V5 .43 .30 .25 .22 .18
V6 .31 .27 .21 .15 .11 .09
V7 .24 .15 .12 .10 .08 .06 .05

 

aFrom Cattell (1952, p. 49).

Thurstone modified Spearman's theory by introducing
the concept of the existence of many factors instead of
solely one as Spearman had postulated. Multifactor analysis
involved the hypothesizing of the existence of several
common factors from a set of correlations. This analysis
(also called tetrad difference) enabled the researcher to

examine many dimensions of the variable and analyze the

relations of the factors to each other at one time, and

39

replaced previous analysis that had to be undertaken on one
factor at a time (Cattell, 1952).

It was necessary for Spearman (1927) in his early
research to develop and refine the new statistical methods
of correlation, 'order,‘ and hierarchy of data. Thurstone
(1935), was also instrumental in advancing traditional
factor analysis; many authors have referred to him as the
father of factor analytic research. He not only helped
find a solution to the problem of attitude measurement, but
he gave impetus and direction to attitude change research
(Ostrom, 1968). It was Thurstone who developed the concept
of 'clustering' which evolved into multiple factor analysis.
The concept of correlation clusters involved positing certain
primary mental abilities with common space between them.
"The area of common overlap in each cluster defines a
primary mental ability." Thurstone hypothesized seven
primary mental abilities, in opposition to Spearman's 'G'
and '5' factors (Chaplin and Krawiec, 1968, pp. 524-525).

Thurstone's support for his theory of Primary Mental
Abilities was based on correlations of various mental tests.
Different tests revealed different 'factor loadings' or
different degrees of closeness to the varying correlational
clusters. The battery of tests could be refined and
improved to correlate more highly with each of the under-

lying factors.

 

 

.-
¢-*-.
"5!
Y

l-t
--|.-
- .—
«Ir.

 

"a .

-.. ‘

u."‘

.C‘ad‘

,.u‘

 

40

Thurstone (1935) introduced into the mainstream of
psychological research the concept of 'mapping' as well as
expanding and developing the previously discovered concept
of hierarchy. Guttman's use of the "mapping sentence" is
deve10ped in detail in Chapter V.

Throughout this research, Thurstone was concerned
with making psychology more rigorous; thus, he attempted to
have his research conform to the scientific method. The
criterion by which he accepted or rejected his studies was
the degree to which "it facilitates the comprehension of a
class of phenomena which can be thought of as examples of a
single construct rather than as individualized events"
(Thurstone, 1947, p. 52).

Thurstone (1947) applied factor analysis to problems
involving measurement of intellectual ability and individual
differences. Unsatisfied with Spearman's interpretation of
intelligence based on 'one' general factor, Thurstone ad-
vanced the concept that intelligence was composed of several
separate factors which appeared in correlational clusters.
Factor analysis was regarded as a general scientific method
to be used in studying individual differences. Thurstone
believed that the main purpose of factor analysis was
indeed for the study of these individual differences. The
assumption underlying factor analysis is that a variety
of behaviors within a certain area are not only related,

but they are at least partially determined by a small number

0 f factors :

41

Factor analysis is useful where basic concepts are
lacking in an area, and experiments have been diffi-
cult to conceive. The new methods have a humble
role. They enable us to make only the crudest first
map of a new domain. But if we have scientific
intuition and sufficient ingenuity, the rough fac-
toral map of a new domain will enable us to proceed
beyond the exploratory factoral stage to the more
direct forms of psychological eXperimentation in the
laboratory (Thurstone, 1947, p. 56).

In strict factor analytic studies, there are no
clearly defined independent or dependent variables; rather,
all the variables of the study are treated in a like manner.
If, however, one desires in a study to predict one variable
from another, a statistical study should be performed
(Thurstone, 1947) .

Factor analysis, as an empirical method, is typ-
ically useful where a researcher is attempting to discover,
explore, and order variables underlying a certain problem.
The factor analyst is especially interested in studying
correlations that he observes, and accounting for these
correlations among the variables "in terms of the smallest
number of factors and with the smallest possible residual
errors" (Thurstone, 1947, p. 60). The process of factor
analysis involves the investigation of a set of variables
to discover if they exhibit, or can be modified to exhibit,
some underlying order that is responsible for producing the
individual differences observed. The variables and data

are then ordered in the form of a matrix. This matrix

formulational analysis of correlations among variables owed

-u

an)

."

-n~

'I

l

42

its beginning to the development of the methods of correla-
tion by Pearson and Galton, Spearman and Hart's technique
of ordering correlations in the form of a hierarchical
matrix, and finally Thurstone's writings stressing the
importance and usefulness of matrix analysis to the fields
of psychology and education. Although it was Spearman and
Hart who first noted matrix analysis, it was left to
Thurstone to popularize the method. As Thurstone wrote in
1947, "the matrix formulation of factor analysis seems to
have been generally accepted, and it has largely replaced
the previous methods of factor analysis" (Thurstone, 1947,
p. VI).

A correlation matrix involves a table of inter-

Correlations, as defined by Thurstone (1947, pp. 1-2) below:
Matrices and determinants involve rectangular
arrangements of numbers. Any rectangular arrange-
ment of numbers is called a matrix, irrespective
of what the numbers mean. If the matrix has m
rows and n columns, the matrix is said to be of
order m x n. In designating the order of a
matrix, it is customary to refer to rows first
and columns second.

Table 5 illustrates one of Thurstone's matrices,
Specifying the product-moment correlations between sixteen
measurements on thirty-two factors, but does not specify
any 'order' or hierarchy of the factors. Future develop-
ments in correlation matrices illustrate the concept of
'ordered' factors in a matrix. When it was found that inter-

correlations such as the ones illustrated in Table 5 exist,

Thurstone considered the matrix to be of 'simple structure.‘

.AHmv .m .nvav mcoumusce Eonmm

 

43

oo.H on. em. vv.l mm.| mo.l HN.I hm.l Nm.l ov.l 00.! Hm.l mH.I NN.I MN.I HH. 0H
on. oo.H mm.l Nm.l mm.l NH.I NN.I vo.l vo.l m©.I 0N.I hm.l NN.I vm.l vH.I mm.l mH
gm. mN.t oo.H NH. gm. HH. Ho. HH. mH. mm. «m. mo. NH. mH. @H.I on. vH
vv.l Nm.l NH. oo.H hm. ow. om. mm. mm. Hm. mm. mm. Nm. mm. v0. mm. MH
mm.l mm.l vm. hm. oo.H mm. mm. om. vm. mm. mm. gm. mm. mm. cm. 00. NH
oo.l NH.I HH. om. hm. oo.H om. 00. we. v0. vm. vm. mm. om. mm. Hm. HH
HN.I NN.I Ho. om. mm. om. oo.H mm. mm. vb. Nm. om. Hm. mv. no. mv. OH
hm.| vo.l HH. mm. om. 00. on. oo.H mm. mm. on. me. Ho. om. Vb. mm. m
Nm.l vo.l 0H. mm. cm. mm. mm. mm. oo.H mm. mm. om. om. Nm. mv. av. m
ov.l m©.I mm. Hm. mm. «0. we. mm. mm. oo.H mm. on. mm. mm. hm. Nb. h
mo.| mN.I cm. mm. mm. cm. Na. 05. mm. mm. oo.H om. om. Nv. NV. mm. o
Hm.l hm.l mo. mm. cm. «0. om. mu. om. mu. om. oo.H mm. Nm. om. mN. m
mH.I NN.I NH. Nm. we. mm. Hm. Ho. om. mm. mm. mm. oo.H Nm. Nm. Nm. v
Nn.l ww.n mH. mm. mm. om. mv. om. Nm. mm. NV. Nm. Nm. oo.H om. om. m
MN.I «H.I 0H.I co. vm. mm. no. es. ov. hm. Nv. om. Nm. om. oo.H om. N
HH. mm.l mm. mm. mm. Hm. mv. mm. av. Nb. mm. mN. Nm. om. om. oo.H H

 

0H mH VH MH NH HH OH m m b m m v m N H

 

m xHHumz coHumHmuuounn.m mqmda

a
‘i‘

.u‘

a
.. v.

u!!!

'
,uuu

 

n;-.

.‘I‘.

l-‘
.

a u“

I...

 

v..

 

44

He suggested that when it was discovered that such a corre-
lation matrix existed between tests, that the tests should
then be administered to several different new populations
for verification and confirmation of the original hypothe-
sized 'simple structure.’ These new and different experi-
mental populations should be selected in ways different
than the way used for selection of the original population.
"If the primary factors are in the nature of basic para-
meters (factors) that are not merely reflections of the
eXperimental conditions or the particular selective con-
ditions, their interpretations should be the same for the
several experimental groups" (Thurstone, 1947, pp. 471-472).

Thurstone's mathematically precise definition of
'simple structure' follows:

Each test may be regarded as a radial vector in a
common-factor space of as many dimensions as there
are common factors in a test battery. The corre-
lation between any pair of tests is the scalar pro-
duct of the test vectors. Since the scalar product
of a pair of vectors is independent of the co-
ordinate system, it follows that the interest
correlations define the co-ordinate system. But

the co-ordinate axes are the scientific categories
in terms of which the tests are to be comprehended.
This is an interesting indeterminacy. One of the
principal problems of factor analysis is to find a
unique set of co-ordinate axes, either orthogonal or
oblique, which shall represent scientifically mean-
ingful categories in terms of which the tests may

be comprehended. This problem has been solved in
terms of what I (Thurstone) have called 'simple
structure' of a trait configuration (Thurstone, 1935,
p. VIII).

According to Ostrom (1968), the period between 1930

and 1950 markedtjmaemergence of attitude theory; at this

45

time, extensive research began into the study of how atti-
tudes were formed and changed. Much of modern theoretical
and empirical explanations of attitude are based on studies
from these two decades. Ostrom credited Thurstone with
being the one theorist most contributive to attitude
research with his solution to the problem of the measure-
ment of attitudes. Those who followed the work of Thurstone
(e.g., Guttman, 1944; Likert, 1932) "accepted this evalua-
tive characteristic . . . Thurstone, and later Likert
(1932) and Guttman (1944), provided a rational methodology
for the measurement of attitudinal affect" (Ostrom, 1968,
Pp. 7~27).

Chaplin and Krawiec (1968) presented a good summa-
tion of what Thurstone and other factor analysts sought to
achieve, and what Jordan is attempting to carry through in
practice in the realm of attitude-behavior research. "When
the smallest number of factors which can account for the
Correlations has been discovered and when the factors have
been identified with their corresponding processes, the
psychologist is in possession of a theoretical description
of the system he is seeking to establish" (p. 527). Vali-
dation of this entire process was based on two contingencies:
(a) the validity of the operations from.which the system
‘was derived, and (b) the psychologist's judgment upon which

the assumptions are based.

 

 

g.

..u

a...

h”

 

~-

46

This dimensional analysis of attitudes was accepted
by those who followed Thurstone. Thurstone believed that
an individual's attitudinal affect depended on the average
of the affective distribution of his personal beliefs.
"Thurstone, and later Likert (1932) and Guttman (1944),
provided a rational methodology for the measurement of
attitudinal affect" (Ostrom, 1968, p. 27). In his intro-
ductory statements concerning factor analysis, Guilford
(1954) described what he perceived as the primary goal

of science:

Science, forever motivated to bring order out of

chaos, to reduce to the simple that which is complex,

wants to know what is the smallest number of concepts

with which one can order and describe adequately the

multiplicity of phenomena that come under its

scrutiny (p. 470).
This task has been undertaken by factor analysts;to discover
the smallest number of variables or dimensions of person-
ality or some other characteristic of human functioning
that will adequately describe that particular functioning.
Through exploration of a specific universe of traits,
factor analysis sought to discover principles of classifi-
cation (Burt, 1966).

Spearman was one of the first to devise a correla-

tion matrix including the concept of order, illustrating
the intercorrelations among variables. Table 6 illustrates

simple proportionality in a correlation matrix. As can be

seen in Table 7, when the same variables are rearranged, the

47

porportionality of the intercorrelations becomes clearer.
The coefficients now run from high to low in every row and
column; this phenomenon Spearman called 'hierarchical order'
(Guilford, 1954).

TABLE 6.-~Intercorre1ations of Six Hypothetical Tests Having

One Common Factor, Illustrating the Condition of
Simple Proportionality in a Correlation Matrix a

 

 

 

a b c d e f

a .40 .10 .45 .30 .35
b .40 .16 .72 .48 .56
c .10 .16 .18 .12 .14
d .45 .72 .18 .54 .63
e .30 .48 .12 .54 .42
f .35 .56 .14 .63 .42

1.60 2.32 .70 2.52 1.86 2.10

aFrom Guilford (1954, p. 473).

TABLE 7.~-Same Intercorrelations as in Table 6 with Variables
Rearranged so as to Show the Proportionality More

 

 

Clearlya

d h f e a c
d .72 .63 .54 .45 .18
b .72 .56 .48 .40 .16
f .63 .56 .42 .35 .14
e .54 .48 .42 .30 .12
a .45 .40 .35 .30 .10
c .18 .16 .14 .12 .10

 

aFrom Guilford (1954, p. 474).

~ v
..v

.a-

48

"Thurstone believed that when simple structure is

achieved in rotation, the factors have psychological meaning.

In other words, simple structure is a principle of order in
psychological nature" (Guilford, 1954, p. 485).

The question of why use factor analysis at all has
been aptly answered by Guilford (1954, p. 522). "The most
defensible reason a psychologist can have for making a
factor analysis is to aim toward the clarification of use-
ful concepts in a domain where adequate concepts are now
lacking."

Cattell (1964) served as a link between Spearman,
Thurstone and other factor analysts who concentrated their
efforts toward understanding intelligence, and the
beginning of the application of factor theory to the area
of personality.

The statementskurCattell concerning linear simplex
theory both explained what Spearman and Thurstone had

achieved with the matrix hierarchy and what Jordan and his

associates have recently encountered in their facet analysis

of attitude-behavior.

Stated now in more detail our postulate for the linear
simplex is that natural relationships may take all
levels of order and complication of mathematical
equations to represent them, but that in the total
pOpulation of relationships (and in the majority

of random samples), there will tend to be a pyramid

or hierarchy of complexity, with the basic linear
relationship as the most common and with increasing
complexity of equation progressively less frequent
(Cattell, 1964, p. 732).

49

Cattell also discussed how errors in the simplex
arise and noted some ways to overcome them; the basic
reasoning he utilized was similar to Guttman's (1944, 1954b)
explanations of why a perfect simplex was not often attained,
and by Jordan (1971a) in his analysis of attitude theory:

On the assumption that the relationship we are
dealing with is linear (since the majority will
be so) any error of estimate from any variable

x to another, y, will be a statistical phenomenon,
due to (a) experimental and sampling error, and
(b) partial determination of the variance of the
second variable by variables other than the first.
The departure of the raw score scaling from the
ideal scaling operates as experimental, instru—
mental error of measurement and this reduces
accuracy of estimate ~~ and the correlation
coefficient ~~ only through source (a) (Cattell,
1964, p. 732).

Cattell was the first to apply factor analytic
methods to an analysis of personality theory, while attemp-
ting to specify its structure. At about the same point in
time, Guttman (1959) began to use facet analysis as his
primary method of investigation. Lingoes and Vandenberg
(1966, p. 2) felt that Guttman's facet theory (1959)
served, in some respects, as "the nonmetric counterpart of
Thurstone's concept of simple structure." The differences
between traditional factor analysis and Guttman's facet
theory can be seen in comparing his system to the factor
system of Cyril Burt. Both Guttman and Burt were attempting
analysis of qualitative data; their techniques were arrived

at independently but bear similarities. Burt's goal was

to factorize the data, while Guttman's primary aim "was to

50

present 'a theory and method of scale construction' by
means of 'quantifying a class of attributes'" (Burt, 1953,
p. 5). Guttman believed that factor analysis was suited
for quantitative data and could not be used when dealing
with qualitative variables (Burt). Guttman's objections
to factor analysis were as follows:

1. Factor analysis is "designed only for quanti-
fiable variables, and is consequently unsuited
for qualitative data."

2. . . . "In order to apply factor analysis, we
must begin by calculating correlation coeffic-
ients, and in the case of qualitative data such
coefficients are bound to be misleading . . .

3. "The principle criterion for scalability is
reproducibility. But factor analysis does not
allow us to reproduce the original data from the
so-called factor-measurements. Hence factor
analysis can never show whether a scale is
perfect or not."

4. "The Spearman-Thurstone approach to factor
analysis is completely linear, and is there-
fore not adequate for analyzing the curvi~
linearities inherent in the scale pattern."

5. "From a scale analysis it can be known what a
factor analysis will show; from a factor
analysis it will usually be difficult, if not
impossible, to know what a scale analysis will
show" (Burt, 1953, pp. lO—ll).

Facet analysis is "a tool for the organization of
ideas" (Foskett, 1963, p. 111). Guttman's techniques serve
as a method for research in social science. Foskett
(p. 111) defined Guttman's facet analysis as "the coordi-
nation of elements from sets which together add up to the
whole content of research projects."

A facet, then, is a set of elements which may combine
with other sets, and 'facets are involved . . . in

almost any scientific endeavour in any field'. The
theory of facets make it possible to design the

51

'universes of content' of research projects econom-
ically, and even to derive new psychological and
sociological concepts (Foskett, 1963, p. 111).

Guttman Methodology

 

Guttman (1954b) based much of his structural research
model on the factor analytic techniques of Spearman and
Thurstone; although Guttman's method is not a factor analytic
theory, he considers factor analysis a predecessor to facet
analysis.

There has been a definite void in discovering and
developing means in the social sciences to quantify qualita-
tive data. In one of Guttman's (1944) early articles, he
sought a recognition of this phenomenon, and also addressed
himself to the task of discovering ways out of this dilemma.
He presented a new approach to this problem by citing some
quantitative methods that could be applied to data and
research that is primarily qualitative in nature.

Guttman began this early article by defining termi-
nology that he would use in building his system. This
terminology has been carried over by Jordan in his own
systematic analysis of attitude-behavior based on the
methodology of Guttman's system.

According to Guttman (1944), "a variable denotes a
set of values; these values may be numerical (quantitative)
or non-numerical (qualitative)." 'Attribute' is used by
Guttman to denote a 'qualitative variable;' these terms

are interchangeable. "The values of an attribute (or of a

52

variable, too, for that matter) may be called its subcate-

gories, or simply categories." A scale is the multivariate

 

frequency distribution of the universe of given variables
in which one can derive a quantitative variable "with which
to categorize the objects such that each attribute is a
simple function of that quantitative variable." These

quantitative variables are called scale variables (pp. 139-

 

140).

Guttman (1944) warned that perfect unidimensional
scales should not be expected in actual, applied research.
A method was developed to test data to find out if it
forms a reasonable estimate of a relatively unidimensional
scale. "The deviation from perfection is measured by the
coefficient of reproducability, which is simply the empirical
relative frequency with which the values of the attributes
do correspond to the proper intervals of a quantitative
variable" (p. 140). If a scale reaches eighty-five percent
or better, Guttman considered this, although not a perfect
scale, an efficient approximation to a 'theoretically
perfect scale.‘ The values for each scale variable are

called either scale scores or scores; the order of the

 

objects in the scale by numerical order of scores is called

their scale order. This 'scale order' is analogous to

 

Spearman and Hart's 'hierarchy' and Thurstone's 'matrix

formulation,‘ and, according to Thurstone, . . . seems to

have been generally accepted, and has largely replaced the

53

previous methods in factor analysis" (Thurstone, 1947,
p. VI). In a more recent work, Guttman referred to this
'scale order' as 'structural theory' (1971).

In devising his method for quantifying qualitative
data, Guttman (1944) relied on what he calls 'the universe
of attributes.’ The effect is directed toward scaling a
universe of attributes, which is a large class of behavior,
and contains 'all' of the attributes under investigation.
"The universe consists of all the attributes that define
the concept" (p. 141). The universe can also be defined
as containing all the attributes that the researcher is
interested in that have some common content; they would be
classified under the same single heading, indicating the
content of the variable class.

An important prOperty of a scalable universe is
that the ordering of persons based on a sample
of items will be essentially that based on the
universe . . . Hence, we are assured that if a
person ranks higher than another in a sample of
items, he will rank higher in the universe of
items. This is an important prOperty of scales,
that from a sample of attributes we can draw
inference about the universe of attributes
(Guttman, 1944, p. 147).

Guttman (1944) distinguished between ordinary
problems of prediction and scaling via the universe of attri-
butes. While in prediction problems, a dependent variable
is to be predicted from the attributes, in the method of
scaling each attribute is predictable from the quantitative

variable. A quantitative variable is derived from the multi—

variate distribution such that each attribute is a simple

54

function of that variable; this procedure is undertaken only
after discovering that a particular universe of attributes
is indeed scalable for a population.

From this method of analysis, can be seen an impor-
tant property of this particular technique, a prOperty
that becomes especially crucial in Jordan's extension of
the Guttman system. Guttman (1944, p. 148) stated that "if
the items have a multivariate distribution that is scalable,
it can easily be seen that no matter what the outside vari-
able may be, the same prediction weights may be given to
the items." Jordan's research involved use of a similar
universe of attributes applied to several different attitude
objects. As suggested by Guttman (1959), Jordan has
enlarged the theory "by letting the groups vary according
to some principle" (p. 319). "The correlation of any
outside variable with the scale scores is precisely the
same as the multiple correlation of that outside variable
with the items in the scale" (Guttman, 1944, p. 418).
Thus, scaling of items having a multivariate distribution
that is scalable, provides an invariant quantification of
the attributes in order to predict any outside variable.
Scale scores from a scalable multivariate distribution,
can serve for almost any prediction purpose defined by the
researcher (Guttman,l944).

Thus, the major difference between scaling and pre-

diction is that in prediction, a variable is predicted from

55

the attributes, while in scaling, an attempt is made to
reproduce the attributes (or variables) from a quanti-
tative variable" (Guttman, 1944).

In a discussion of the relativity of scales,
Guttman (1944) gave an analysis of deviant or non-scale
types. If these latter types are too numerous, thereby
keeping the coefficient of reproducability below the
eighty-five percent criterion Guttman has established, a
unidimensional scale cannot be said to exist. The reason
scales are derived is because there is a certain degree of
uniformity of experience for the population being tested in
order that the attributes have a similar meaning to the
different individuals taking the test. Guttman added that
the individuals who deviated from the scale analysis may
be useful for some type of in-depth case study analysis.

A perfect scale order is dependent on only one
component, that of the rank order underlying the attributes.
One can deduce an individual's attitude or behavior on
every item in the universe of attributes being evaluated,
by observing his scale rank. The totality of behavior and
the interrelationships between items can be measured by a
single variable, the scale rank (Guttman, 1954b).

Guttman (1954b) divided factor analysis into two
basic types: (a) common factors, the previous approach used
by Spearman, Thurstone, and others; and (b) the method of

order-factors, Guttman's own approach. Guttman did not feel

56

that these two approaches were totally different, but con-
sidered earlier factor analytic techniques as predecessors
to his theory.

Shortly after its inception, various factor analytic
methods began to converge toward a design and theory for
measuring mental abilities. Guttman (1958) described these
convergences and other relevant issues in his article,

"What Lies Ahead for Factor Analysis." The work of Thompson,
Thurstone, and Spearman centered around this search. Much
of this theorizing was based on Spearman's concept of 'G'
and its association with Thurstone's concept of 'simple
structure.‘ Similar designs were developed by El Koussy,
Guilford, and Guttman. El Koussy's design involved the
study of ability and physical space. Guilford devised a
facet theory based on many of the studies involving Thurs-
tone's concept of simple structure. In his extensive study
of the techniques of multivariate analysis used in the be-
havioral sciences, Guttman arrived at a similar design
(Guttman, 1958).

Both Spearman and Thurstone were interested in psy-
chological theories of mental abilities and the more abstract
statistical and algebraic theory. Thurstone stressed the
need for the algebra and statistics of factor analysis to
be used in the investigation of psychological conceptions.
Although his development of multiple-factor analysis

appeared largely mathematical, Guttman (1958) stated that it

57

was primarily motivated by psychological considerations.
Thurstone's multiple-factor theory grew out of Spearman's
single-factor theory. By dividing Spearman's single-factor
theory into several common-factors, Thurstone enumerated

a multiple theory for the intercorrelations of mental
abilities. Thurstone believed that the number of factors
responsible for mental ability, should be relatively few in
number compared to the large number possible for mental
abilities or tests of mental ability. In opposition to
Spearman, however, simple structure posits that not all
common factors are involved in all the diverse mental
abilities (Guttman, 1958).

Guttman (1958) continued to discuss the convergences
of factor analytic theories by noting similarities in the
work of E1 Koussy, Burt, and Spearman. Close parallels can
be found in E1 Koussy's space abilities research, Spearman's
psychological concepts involved with his single-factor
theory, and Burt's conception of hierarchical levels of
factors.

Guilford, on the other hand, drew upon the vast
number of common-factors that had already been discovered
by others as related to intellectual ability, and compiled
a new, more complete listing of factors related to intelli-
gence. Interested in finding psychological meaning for
these factors that could be structured, Guilford devised a

three-faceted scheme for intellectual ability. In this

58

scheme, he distinguished five types of intellect (memory,
cognition, convergent thinking, divergent thinking, and
evaluation) which were further divided into three types of
content (figural, structural, and conceptual). This yielded
5 x 3 fifteen possible combinations of intellect and
content. A third facet, type of thing, divided these
fifteen elements into six additional subclasses (fundaments,
classes, relations, patterns of systems, problems and impli-
cations). Thus, from three facets (intellect, content, and
thing), 15 x 6 possible combinations were possible. Guttman
stated that although these ninety common-factors may not be
complete or correct in every detail, it was the design of
the theory that would be so crucial to the future direction
of studies in factor analysis (Guttman, 1958).
These above designs (E1 Koussy and Guilford) were

referred to by mathematicians as Cartesian Products of Sets.

If I is a set of intellectual abilities, if C is

the set of three types of content, and T is the set

of six types of things, then by the Cartesian product

ICT is meant the set of ordered triples, say of the

form ict, where i is an element of I, c is an element

of C, and t is an element of T. Each set in a

Cartesian product is what Fisher calls a 'factor'

for his design of experiments, and an element of such

a 'factor' is what he calls 'level of a factor'.

Since this use of the word 'factor' is radically

different from that of Spearman and Thurstone we

have proposed that the word facet be used instead of

Fisher's. A facet is nothing but a set involved in

a Cartesian product. I, C, and T are the three

facets of the Cartesian product ICT (Guttman, 1958,
p. 508).

59

"Following Spearman, should we regard 'qualitative
law' or a facet design as being a set of instructions for
a test constructor as to what kinds of items he should make
up for tests, then Guilford's ICT design provides instruc~
(zions for ninety varieties of tests" (Guttman, 1958, p. 508).
thter a test was constructed and administered, the task
vnould be to explain why a certain subject scored as he did
111 terms of the facets of the test design.

Guttman's (1958) radex theory of mental abilities
lxmgan with Spearman's two general facets of complexity,
'Imalations' and 'fundaments.‘ By varying the level of
conqplexity while holding content constant, a simplex corre-
Lation matrix could be obtained. Similarly, by varying
mmntent and holding level of complexity constant, it was
possible to obtain a circumplex correlation matrix. Fur-
ther, 'kind of complexity' could be expanded through use of
E1 Koussy's and Guilford's facets.

Guttman (1958) suggested that after a design had
beconma accepted for one area of endeavor (such as mental
abilities), it should be possible to apply the design
through modification to other areas of concern. In dis-
coveringthe facet design for mental abilities, one also
discOvered the definition of mental abilities; their content
is identical. The old definition of intelligence, "what an
intelligence test measures, was facetless, and therefore

quite useless for empirical study" (Guttman, 1958).

60

If the design was incomplete, additional facets

and/or modification of the facets present might be necessary

(Guttman, 1958).
There is a great deal of work to be done on the
purely conceptual level of designing a cleanly
faceted system, or a definition of mental abilities.
Guilford's, El Koussy's, and my own (Guttman's) two
primitive facets by no means include all of those
indicated by Spearman. A good deal of fruitful
work in the future may go in the direction of simply
deciding upon what one wants to mean by mental
abilities in terms of facets. We have seen how
major concepts of our great predecessors are retained

in this reformulation, but retained in what may be
a cleaner and less ambiguous form (Guttman, 1958,

p. 514).

Guttman (1954b) described his 'radex theory,‘
Whixch he designated to indicate 'radial expansion of a com-
plexity,‘ as "a set of variables whose intercorrelations
COrLform to the general order pattern prescribed by the new
thenory" (p. 260). A set of variables that possesses a
Sinuple order of complexity and can be arranged in a simple
rank order from least complex to most complex, is called a
“iinuplex.' The variables contained within the simplex
dififear in the degree of their complexity. This holds for
tests of similar kind; for example, in a group of tests

meelsuring numerical ability, addition, subtraction, multi-

pliSIation, and division differ largely in degree of complex-

ity, from less complex to more complex.
In opposition to this simplex design, was what
Guttman referred to as a 'circumplex.’ The circumplex

contains a 'circular order of complexity,‘ in which the

61

order is not from least complex to most complex, but rather
has a circular order. Tests which contain the same degree

of complexity will differ only in the kind of ability that

they define. "Our empirical data will testify that differ-
ent abilities such as verbal, numerical, reasoning, etc.,
(ho tend to have such an order among themselves" (Guttman,
.1954b, p. 260).

Since the simplex analysis is so central to
(Hardan's (l97la) analysis of attitude-behavior, a brief
rexziew of the foundations that this theory rests upon is
immxortant. Spearman's early studies were based on postu-
laILing hierarchies for the relationship between two tests;
these were arranged in such an order that the correlations
betnfleen the tests decreased the further down in the hierarchy
they were located. If you locate a number in the upper
Jfafflz corner of the table, moving anywhere to the right or
dOVnn. in the table, will result in the correlations tapering
Off? (Table 8). Guttman (1954b), however, stated that
attErnpts to apply this 'hierarchy hypothesis' met with
faiJLIJre, and the word 'hierarchy' disappeared from the
liteBirature. Guttman, especially in his analysis of simplex
datii. is attempting to revive usage of this concept.

Spearman's theory was based upon the assumption that
one"‘common-factor held the hierarchy together; when one

factor was postulated it was found that the hierarchical

relationship could not be produced. Other theorists then

nn

 

 

.7
‘-

 

 

62

TABLE 8.~~Test Intercorrelations for a Hypothetical Single-
Common-Factor Structurea

 

 

 

Test Comgggagigtor t1 t2 t2;f t4 t5
.9 .7 .5 .3 .1
t1 .9 (.81) .63 .45 .27 .09
t2 .7 .63 (.49) .35 .21 .07
t3 .5 .45 .35 (.25) .15 .05
t4 .3 .27 .21 .15 (.09) .03
t5 .1 .09 .07 .15 .03 (.01)

 

aFrom Guttman (1954b, p. 263).

postulated two, three, four, and even more common factors
as central to leading to the hierarchical order; however,
even these failed to lead to reproduction of the hierarchy.
Guttman at first postulated that multiple common-factors
are needed to explain the intercorrelations among the
variables; this he stated proved to be an empirical failure,
although it was mathematically accurate. Through modifi-
cation of this theory, he has arrived at an alternative
explanation of the system, that being the simplex. A
theory closely related to the simplex is that of the circum-
plex, and both these theories are encompassed in the more
comprehensive theory, the radex. This new approach encom—
passed much from the older theories and also eliminated many
of the older approaches (Guttman, 1954b).

Guttman (1954b) considered the simplex, a facet

theory, as being a viable alternative single-factor theory

63

to the older factor analytic technique, that of Spearman's
single-common-factor hypothesis, while not utilizing an
m-common-factor theory. The alternative single-factor
theory involved the use of scaling and scale analysis.

By giving semantic meaning to a rank order among quantita-
tive variables, the hierarchy previously abandoned by those
seeking multiple-common-factors, can be revived to serve
as a useful means of analysis. Guttman (Guttman & Schles-
inger, 1966; 1967) felt that this new facet analysis was
feasible for mental tests, both theoretically and in
applied research through use of the concept of complexity
as the basis for comparing different variables. This was
one juncture where Guttman replaced the older factor
analysis with his own facet techniques:

. t

Suppose we are given n tests t n

,t..
which differ only on a single $0mp1exity
factor . . . Test t is the least complex. Test
t is next; it requires everything t does, and
m re. Similarly, t is more complex than t2,
requiring everything t does and more. In
this case, t is clearly also more complex than
t1. In general, t. + l is more complex than t.,
and hence requiresjwhat all proceeding tests
require, plus something more. Let g denote the
total complexity factor, of which all tests are
composed in various degrees. Thus, 9 is like an
additional test beyond the most complex given test
tn (Guttman, 1954b, p. 269).

This initial ordering was essential for Guttman's
scaling technique; with Spearman's hierarchy no such ordering
was required. As previously stated, Spearman's theory pro-

duced a matrix (see Table 8) that descends as one goes down

64

or to the right of the upper left corner. The results of
Guttman's theory can be seen to be quite different from
what Spearman had found. Table 9 illustrates how Guttman's
theory lead to a matrix in which the largest correlations
all lie along the central diagonal, and taper off as one
goes to the upper right and lower left of the table
(Guttman, 1954b).

TABLE 9.~~Test Intercorrelations for a Hypothetical Equally-
Spaced, Perfect Simplexa

 

 

 

 

t t t t t
Test Complexity 1 2 3 4 5
.07776 .1296 .216 .36 .6
t1 .07776 1.0 .6 .36 .216 .1296
t2 .1296 .6 1.0 .6 .36 .216
t3 .216 .36 .6 1.0 .6 ..36
t4 .36 .216 .36 .6 1.0 .6
t5 .6 .1296 .216 .36 .6 1.0
Total 2.3056 2.7760 2.9200 2.7760 1.3056

 

aFrom Guttman (1954b, p. 271).

Not all matrices are as equally-spaced as the one
shown in Table 9, "Test Intercorrelations for a Hypotheti-
cal, Equally-Spaced, Perfect Simplex;" in practice few
reach this level of perfection. Table 10 illustrates the
more common results, where the data do not form a perfect,
equally-spaced somplex, but form a "Hypothetical, Nonequally-
Spaced, Perfect Simplex." The tests are not equally-spaced

in their complexity, but they still maintain the pattern

65

of descending scores as one proceeds further away from the

central diagonal (Guttman, 1954b).

TABLE 10.~-Test Intercorrelations for a Hypothetical, Non-
equally-Spaced, Perfect Simplexa

 

 

 

 

Test Complexity t1 t2 t3 t4 t5
Loading .10 .12 .30 .85 .90
t1 .10 1.00 .83 .33 .12 .11
t2 .12 .83 1.00 .40 .14 .13
t3 .30 .33 .40 1.00 .35 .33
t4 .85 .12 .14 .35 1.00 .94
t5 .90 .11 .13 .33 .94 1.00
Total 2.39 2.50 2.41 2.55 2.51

 

aFrom Guttman (1954b, p. 272).

"A set of tests whose observed intercorrelations
satisfy the stated conditions of the hierarchy, will be
said to form a perfect simplex. They have a simple order
of complexity" (Guttman, 1944, p. 271). "In a perfect
scale, each item is a perfect function of a single rank
order of respondents" (Guttman, l953b,p. 2).

Guttman (1954b) regarded Spearman's hierarchy as
one in which the included tests have the same level of
complexity. It was based on "the relative size of the
communalities, or the saturations with the single-common~
factor" (p. 318).

Thurstone's (1947) conception of simple structure

was more closely related to Guttman's notion of circumplex

66

than to his notion of simplex. The 'simple structure
concept,‘ an important contribution of traditional factor
analysis, involved a situation in which the common-factors
did not correlate at all with each other, as was the case
with quantitative variables when using the circumplex.
Guttman (1954b) stressed the point that new theories were
based to a large extent on the contributions of previous
theories.

In visualizing Guttman's simplex, it becomes
increasingly clear just how important previous factor
analytic theory has been to his new theoretical developments
and modifications, evolving into a new facet theory approach.

When Godfrey Thomson demonstrated that Spearman's
type of hierarchy could tend to be accounted for
by a theory of random sampling of 'bonds' in the
mind, Spearman objected on the grounds that mental
activity was certainly not random. Our (Guttman's)
new theory is essentially one of 'ordered—bonds' in
the mind; Thomson's bonds in a sense remain, and
Spearman's objection to randomness is sustained
(but Spearman's hierarchy is displaced from its
previous central importance) (Guttman, 1954b,

pp. 345-346).

Although Guttman (1954b) was somewhat critical of
Thurstone's methodology, he saw more similarity between
Thurstone and his facet theory analysis, than with other
factor analytic theories. "The emphasis that Thurstone
makes on patterns of zero factor loadings is reached in the
additive forms of both the circumplex and the simplex. Had

the notion of a simplex order been available before, one

might have arrived at the radex theory via Thurstone's

67

approach" (p. 346). "We can now see that in a radex with
empirically distinguishable simplexes, centroids of the
simplexes will tend to be the reference axes of a Thurstone-
type analysis. Thus, the number of common-factors found
previously will tend to correSpond to the number of sim-
plexes employed" (p. 347).

Guttman (1953a) stated that his methods have
"demonstrated that image theory is related to common-factor
theory but has greater generality than common-factor theory,
being able to deal with structures other than those des-
cribed in a Spearman-Thurstone factor space" (p. 277).
Guttman felt that image analysis encompassed the common-
factor analysis propunded by Spearman and Thurstone.

Cluster analysis, propounded by Tryon, Cattell,
etc., resembled both circumplex and simplex analysis.
Although Tryon's (Guttman, 1954b) 'correlation profile'
technique fit well in a short simplex, in the more lengthy
tables, "the earlier and later parts will seem to form
separate clusters because of the smaller correlations in the
northeast and southwest corners of the correlation table"
(p. 347). The research performed by Cattell came close to
the simplex proposed by Guttman, but Cattell failed to make
a distinction between kind and degree.

In a 1954-1955 article, Guttman mentioned possible
extensions of his simplex, circumplex, and radex theories

into the realm of attitude research. "One of the most

68

surprising and profound prOperties of a perfect scale is that
while on the surface it represents but a simple ranking of
people ~~ more deeply it reveals a whole series of under-
lying components of the attitude" (p. 400). Guttman con-
tinued to cite the need in attitude research for the scaling
of attitudes along a simplex from 'most complex' to 'least
complex' as a method of improving psychological investi~
gation and computing techniques, over the previous analytic
approaches.

Guttman was of the belief that there was definitely
a structural analysis underlying content areas and that it
was the job of the social scientist to seek out these
underlying structures, where they existed. The method he
advocated for discovering these foundations was facet
analysis of the sub-universes. Certain subuniverses were
statistically closer to other subuniverses, while some were
more distant. These could be aligned in the hierarchical
simplex in order that they specify the empirical correla-
tions underlying the definitions. The more exact that this
analysis became, the better the frame-work was. "Comprehen-
sion of the multivariate system of the universe can lead to
larger theories of relations with other universes, and thus
to more and more perfect multiple correlations for each
variety of behavior separately" (Guttman, 1959, p. 318).

Cattell (1964) in regard to simplex analysis, related

that the concept of simplex had been used widely in

69

scientific endeavors and classically expressed by Newton as
"natura est simplex.‘ In more modern usage this concept
involved choosing the model (or theory) that gave the best
fit and was at the same time simpler in terms of data rela-
tions, mathematics, and logic. Cattell stated that the aim
of simplex analysis was to find scales that would maximize
prediction of certain statistical relationships.

Borgatta (1958) noted that a primary advantage of
data arranged along a simplex design, was that as one studies
the scores down the central diagonal, one can View the
transition in meaning of the items "as they are saturated in
two different contents" (p. 525). If the items chosen were
relatively independent of one another, good factor defini-
tions would not result; if all of the items were totally
independent of the other items, common factors would not be
found and all the intercorrelations would be zero.

According to Foa (1958) one of the most important
characteristics of facet analysis was that it contributed
to the understanding of the structural pattern of contents;
facet theory attempted to show that changes in a given
area depended on changes taking place in the areas imme-
diately neighboring it. It attempted to provide a descrip-
tion of the phenomena that take place and the psychological
dimensions underlying them.

Foa (1958) discussed model building as the first

step in facet analysis; selection of the facets should be

70

made so that they are relatively independent of one another
and yet logically consistent. This he referred to as the
"principle of logical independence of the facets" (p. 230).
Each combination of facet categories should make logical
sense. Foa gave an example using two facets, i.e., sex
(male/female) and motherhood/non-motherhood. He stated that
these two facets were not logically independent because the
combination of male/motherhood was not logically possible.
Although it is important to have the facets of a
scale logically independent, more proof that the facets are
indeed the appropriate facets for the design is necessary.
"A more conclusive indication of the adequacy of choice
consists in finding a certain parallelism between the pattern
of conceptual contiguity and the pattern of statistical
dependence" (Foa, 1958, p. 230). In other words, parts of
similar composition should be related more clearly than
parts of dissimilar composition. When this does not occur,
and if the parts of dissimilar facet composition are more
closely related than are those of similar composition, there
is a good chance that there is some additional, underlying
facet in the model that has escaped notice and has been
omitted from the scale. "These considerations can be
summarized in the principle of contiguity stating that:
conceptual contiguity is a necessary condition for statis-

tical dependence" (p. 230).

71

Guttman considered himself a disciple of both Spear-
man and Thurstone. In tracing the concept of simple struc-
ture, Guttman credited both Spearman and Thurstone for
developing the primary notions leading to the concept of
simplex. "It is a historical fact regarding the evolution
of ideas that the origins of property, facet, and aspect
analysis happened to lie in the early quantitative factor
analytic developments, though they could have been developed
quite independently as a purely logical system" (Guttman,
1966, p. 444). Guttman noted that when Spearman first began
his work, that eventually lead into 'order' analysis, the
statistical means of multivariate analysis were not yet
developed. This impeded the development of the statistical
approaches that Guttman was able to facilitate soon after
the new techniques became available.

Guilford (1954) suggested that patterns of observed
correlations could be predicted if the underlying facet
design of the test is understood. Later, Guttman developed
the theoretical, methodological concept of 'radex' to
explain the relationships of ordered facets. Thus, Guttman's
concept of 'radex' came from Guilford's discovery that a
test with a profile in which there is only one level reversal
will have a higher correlation than tests with more than one
levelreversa1.(Guttman, 1958, p. 512). "It is this kind of
property (radex analysis) that can lead to parsimonious

but highly successful prediction of external criteria"

~v

 

. u
«I.

q .
A.»

..L

 

72

(p. 512). Guttman (1954b), through simplex and circumplex
analysis, has shown how this 'law of parsimony' can be
applied to a variety of empirical testing of different
content areas (Ben-Sira & Guttman, 1971; Guttman &
Schlesinger, 1966).

In attempting to apply his theory to practical
problems of society, Guttman (1944, 1954b, 1959) employed
the theory of radex and the 'law of parsimony' to the
empirical domains of intelligence and racial attitudes.

In the late 1960's and early 1970's, further application of
this theory was undertaken as Jordan and others extended the
simplex theory to other attitude content areas such as
mental retardation (Jordan, 1970; Morin, 1969), racial-
ethnic attitudes (Hamersma, 1969; Dell Orto, 1970;
Frechette, 1970; Jordan, 1971b; Jordan l972a;Bray, 1972),
and drug addiction (Kaple, 1971; Nicholson, 1972).

Green (1954) stated that in 1944, Guttman provided
a new departure for the analysis of attitude research.
Through the scaling of monotone attitude items, Guttman
provided a demonstration of cumulative scaling. For an
individual taking the scale, if he responds positively to one
level on the scale, he should also respond positively to
the scale items above that particular level (see Table 11).

There were previous studies using similar scales
(Bogardus, 1925; May and Hartshorne, 1926), but as a general

method of scaling, they were in the main ignored. "It was

73

TABLE 11.~~Scalogram Diagram for Three Dichotomous Itemsa

 

 

Items
3 2 1
Individuals 1 + +
(scale types) 2 _ +
3 - - +
4 _ _ _

 

aFrom B. F. Green (1954, p. 354).
+ indicates a positive response

- indicates a negative response

Guttman who advocated the use of such cumulative items as

a basis for a formal method of scaling. His development of
scalogram analysis stimulated widespread interest in the
method" (Green, 1954, p. 353).

Guttman (1944) believed that any system involving
the use of scalogram analysis techniques had the advantage
over other systems where it was only possible to think of
the variables in terms of clusters of unknown factors.

Using the system of scalogram analysis, the variables could
be placed along a single continuum of interest.

In a more thorough analysis of scalogram techniques,
Guttman (1950a) stated the basic theory underlying his
system. Scalogram analysis is utilized to answer two basic
questions: "(a) the determination of unidimensionality
and (b) the determination of a fixed point of reference
along such a single dimension" (p. 46). For measurements to

be meaningful they must lie along one plane of analysis in

,-

'-

c.

. .

‘1

a.

Pl!

74

which a fixed point can be determined. Scale analysis (or
scalogram analysis) serves as a method for locating quali-
tative data along such a single plane, thus making it uni-
dimensional. The intensity function gives the data the
single anchoring point for the analysis.

Scalogram research is based on the following defini-
tion of a scale: "It requires that each person's responses
shall be reproducible from his rank alone. Each item shall
be a simple function of the person's rank" (Guttman, 1950b,
p. 62). When this requirement is fulfilled the two other
prominant definitions of scales are also fulfilled. That
is, "within each item, if one response is higher than
another, then all people in the higher category must have
higher scale ranks than those in the lower category" and
the second definition that "a person with a higher rank
than another person is just as high or higher on every item
than the other person" (p. 62).

Scalogram theory asserts that for any particular

sample of questions taken from the universe of scaled items
will result in the same rank ordering of individuals.
Scale analysis does not attempt to define the content of an
area, but rather assumes the content is already defined and
tests for representation of the variable in question
(Guttman, 1950b).

Guttman (1950b) stated that scale analysis can be

used for any universe of qualitative attributes, in any

.-

In.

 

H

a

75

science, and through any method of observation. It is
especially suited for areas such as attitude measurement and
public Opinion (Ben-Sira & Guttman, 1971) where methods for
determining methodologically significant data have been
largely lacking. Scale analysis provides means of selecting
items in a large sample of questions that are representative
of the particular unidimensionality of the scale constructed.
That a rank order of subjects can be established for
material that is qualitative in nature is especially signifi-
cant. By means of scalogram analysis, qualitative data can
be interpreted through quantitative means; the qualitative
variable is given quantitative significance "such that each
attribute in the universe of attributes is a simple function
of that quantitative variable" (p. 88). In essence, any
form of data derived from observation (questionnaires,
participant observation, interviews) can be subjected to
scalogram techniques.

Scale analysis is suited for single-factor analysis
of qualitative data, whereas factor analysis is more useful
for the study of quantitative data. Factor analysis cannot
test the scalability of qualitative variables. In the
Spearman-Thurstone tradition of factor analysis, it was not
designed to detect if a set of dichotomies forms a scale.

Jordan has been using facet theory in his research
at Michigan State University; at first, starting out with

faceted definitions of the universe of attitudes toward the

76

mentally retarded, secondly, eXpanding this to include
attitudes toward the mentally ill, and finally expanding it
to other 'personal' attitude object groups, such as race,
drug users, and the war disabled. Currently, he is exploring
the model to study attitude-behavior toward 'conceptual'
attitude objects such as: the role of women, technical
education, educational change, the environment, and pOpu~
lation problems.

Guttman (l953a) stated that his methods have
"demonstrated that image theory is related to common-factor
theory but has greater generality than common-factor theory,
being able to deal with structures other than those des-
cribed in a Spearman-Thurstone factor space" (p. 277).
Guttman felt that image analysis encoumpasses the common-
factor analysis propounded by Spearman and Thurstone.

It was with the work of Jordan and his colleagues
that Guttman's structurally precise facet technique and con~
ceptions concerning attitudes, were fused together into an
evolving theory of attitude-behavior structure and measure-
ment, with its implicit implications for attitude-behavior
change.

Theoretical Developments in Attitude-
Behavior Research

 

As was stated earlier in this paper, Jordan's
attitude research involves a combination of two separate

streams of thought in psychology. His work involves a

77

combination of rigorous semantic and methodological analysis
encompassed within an empirical, theoretical, and descriptive
tradition. In synthesizing these two approaches, he is
attempting to explain attitude-behaviors through a facet-
design approach. The semantic-structural segment of this
analysis involves use of the facet theory of Guttman and his
predecessors; the empirical-descriptive analysis he uses is
discussed in the following section.

"Many authors, past and present, would agree that the
interaction of individual minds produces a common manner of
thinking, feeling, and willing, different from that of 1
single minds in isolation, or from the mere summations of
minds" (Allport, 1954, p. 31). This distinction between
thinking, feeling, and willing, is crucial in the analysis
of Jordan's attitude-behavior research.

The beginning of this three-phase analysis of
attitude, that of cognition, affection, and conation, can
be traced to the time of the early Greek philosophers.
"Plato . . . conceived of the mind as made up of three
facilities or 'institutions.‘ To Plato abdomen was the
seat of emotions or feeling; the breast the seat of striving
and action; the head the seat of reason and thought"
(Allport, 1954, p. 19). Allport (1954) stated that this
trichotomy has persisted to the present time; the terms
have been modified by different writers, but the meaning

has been retained. Plato had classified 'mind' into:

78

affection (feeling), conation (striving), and cognition
(thought).

Fechner was one of the first psychologists to use
attitude scaling in the scientific framework. As Chaplin
and Krawiec (1968) stated, his "methods have stood the test
of time to become fundamental procedures in psychophysical

:measurements, mental testing, and attitude scaling . . ."
up. 40). Titchener, a student of Wundt, brought the new,
rixgorous experimentation to the American continent from
ELLrOpe. He analyzed consciousness into three basic
elements ~ sensations, images, and affective states. In
general, Wundt's followers stressed the importance of
feelings in analyzing attitudes. Clarke, a student of
Titzchener, broke attitude into three conscious phenomena:
innéigyery, sensation, and affection. This 'new' psychology

of structure concentrated on the Operations and function

'V'i—tzhin' the organism alone.

It was necessary for another aspect, that of inter-
action between the subject and the object, to be developed
136213(are the work of attitude-behavior research could mature.
(PEIEE ZBritish act psychologists, James, Ward, Dewey, Angell,
611.51 Carr were the first to study the interaction between
t11433 observer and the observed. Later, in America, Cooley

‘51 Mead took up the act psychologists' interest in the

s
nub>ject~object relationship, and developed the early group

79

psychology approach by studying the interactive system

between two or more subjects.

Thomas and Znaniecki, in a 1918 publication,

defined attitude as: "individual mental processes which

determine both the actual and potential responses of each

person in the social world." Further, they defined a

.socdal value as "any datum having an empirical context
éuzcessible to the members of some social group and a meaning

with regard to which it is or may be an object of activity"

(Allport, 1954, p. 45).
Ward (1920) developed an elaborate system centered

artatnid the relation of the active subject to the object.

‘He? (livided the subject matter of psychology into cognition,

fee 1 ing, and conation .
A predecessor of Ward, James Sully (1892), attempted

tx’ cleassify and divide various mental states into distinc-

'tl"€e modes of human expression. The three modes that he

<20r1<2Gaptualized were feeling: knowing, and willing. Sully

gta\;€2 an example of these three states; it is clear in this

e
Xelrnl]ple that Sully's analysis involved all three character-

feeling, and action. Many

is; - _ . .
t53L<cs of attitude: cognition,
Ps: . . . .
IBPCEIJOlogists follow1ng him With more experimental-type

WC)
151<L fail to include the conative element. Even when the

re:
SQarcher includes this element, the actual behavior of a

$111
Ii<ect toward an object is selfom enumerated. Although

311
3L:1qy's work was purely descriptive, he did contribute to

80

empirical research by emphasizing the importance of recog-
nizing the three separate and distinct elements of attitude:

To illustrate the difference between these modes of
mental manifestation, we may select almost any example
of a familiar mental experience. For instance, I see
an apple on a tree. I may be affected by the beauty
of its colour glowing in the midst of its cool green
surroundings. Such a mental state of delightful
admiration would be properly described as a feeling
or affective state . . . 0r, again, if I happen to

be a connoisseur of apples, my mind may be stimu-
lated by the sight of the object to note its

peculiar variety to which it belongs. Such a
direction of mental activity would come under the
head of knowing, cognitive process or intellec-

tion . . . And, lastly, if I happen to be hot and
thirsty the sight of the apply may very likely

excite a desire to pluck and eat it and prompt the
corresponding actions. And in this case what goes

on in my mind would be a case of willing, volition,
or conation. All mental processes lie under one

of these heads (Sully, 1892, Vol. I, p. 60).

Sully's (1892) classification of mental processes
into three separate categories can be compared to McGuire's
(1969) more recent classification of these three elements
as basic to attitudes. Under Intellection (McGuire-cogni-
tion), Sully included such things as: perception, memory,
and thought. Under Feelings (McGuire-affection), he included:
feelings (i.e., pleasure and pain), emotions (fear, sympathy,
intellectual or logical feelings, aesthetic sentiment,
ethical or moral sentiment). Finally. under Conation or
Volition (McGuire-conation), he included: voluntary move-
ments, desire, and habit.

Ward (1920, p. 56) anticipated Jordan's analysis
of attitude-behaviors. He included the three elements of

attitude, which he referred to as the three 'commonly

81

accepted constituents of experience,’ in what appeared to be
an elementary design of Jordan's mapping sentence, using the

crucial subject-object relationship:

(1) non-voluntary attending
to changes in the sensory Presentation
continuum; ‘= of sensory
(Cognition) 5

(2) being, in consequence,
A SUBJECT either pleased or pained; OBJECTS
’ (Feeling)

(3) by voluntary attention or
'innervation' producing

 

. Present tion
changes in the motor- = a
. of motor
continuum.
(Conation)

Ward (1920) postulated that the important contribution of
this system was that it "traversed both the old bi-partite
and the prevailing tripartite schemes, viz. that between
the subject, on the one hand, as acting and feeling, and
the objects of this activity on the other" (p. 56).
McDougall also used the subject-object-activity continuum in
more experimental research (Boring, 1950).

Allport (1954) indicated that 'attitude' has had
more than one meaning through its historical perspective.
There have been two prevalent, and at time opposing defini-
tions of attitude: one meaning, stressing mental attitude
and one, motor attitude or in Jordan's terms, one stressing
"predisposition" and the second stressing "overt action" or
behavior. The definition of attitude as mental activity

preceded the one in which attitude was given motor

82

significance. The earlier definition perceived attitude as
"a subjective or mental state of preparation for action."
Spencer, one of the first psychologists to use the term

'attitude' wrote in First Principles, published in 1862:

 

Arriving at correct judgments on disputed questions,
much depends on the attitude of mind we preserve
while listening to, or taking part, in the contro-
versy; and for the preservation of a right attitude
it is needful that we should learn how true and

yet how untrue, are average human beliefs (Allport,
1954, p. 43).

It was not until somewhat later that the conception
of attitude as an activity began to appear. In the late
Nineteenth Century, N. Lange (1888) spoke of attitude
referring to perception as resulting from muscular prep-
aration or 'set.' Concurrently with this, Munsterberg
(1889) and Féré (1890) developed theories that centered
around motor and action type attitudes. Wundt and Titchener's
contributions occurred shortly after this. "As a result of
the Wurzburg work (primarily Titchener and Wundt) most
psychologists came to accept attitudes, but not all believed
them to be impalpable and irreducible mental elements. The
students of Titchener felt that attitudes were feelings.
Prior to this, the Wurzburg school attempted to use intro—
spection to resolve the meaning of attitude" (Allport, 1954,
pp. 43-44).

Attitudes were really not well represented or

accepted until the time of Freud. He endowed attitudes with

vitality, equating them with longing, hatred and love, with

83

passion and prejudice, in short, with the onrushing stream
of unconscious life" (Allport, 1954, p. 44). With this
analysis, plus the beginning of experimental investigation

of attitude, attitude received a permanent place in the field
of psychology. Credit must be given to Thomas and Znaniecki
(1918) for giving attitude a permanent place in the field

of sociology and sociological analysis (Allport, 1954).

In the early part of the Twentieth Century, research-
ers began propounding definitions of attitude that in the
main included both the thinking processes of attitude and its
motor aspects. Allport (1954, p. 45) cited the following
as typical definitions of attitude at about this time:

Attitude = the specific mental disposition toward an
incoming (or arising) experience, whereby that ex-
perience is modified; or, a condition of readiness for
a certain type of activity (Warren, 1934).

An attitude is a mental disposition of the human
individual to act for or against a definite object
(Droba, 1933).

An attitude is a mental or neural state of readiness,
organized through experience, exerting a directive

or dynamic influence upon the individual's response

to all objects and situations with which it is re-
lated (Allport, 1935).

The Conception of Attitude as
Attitude-Behavior

 

 

Hamersma (1969) stated that a major problem with
measures of social attitude was that they were not consistent
with actual overt behavior. In his review of the literature

of studies concerned with racial attitudes, he noted that

84

most studies had dealt with 'stereotypic' attitudes and
rarely deal with an individual's actual behavior in real-
world situations.

Guttman (1950a) has operationally defined attitude
as "a delimited totality of behavior with respect to some-
thing" (p. 51). When attitudes are thus defined, items in
an attitude scale can be constructed to yield information
concerned with an individual's actual overt behavior and
eXperience; this should link 'attitudes' and 'behaviors,’
eliminating much of the criticism previously leveled at
attitude studies for failing to do this.

Jordan (l971a) stated that there were two views
that permeated the literature on attitudes and attitude
research. The first, defined attitude as "predisposition
to behavior;" the second emphasized attitude as 'behavior.'
Jordan's definition involved use of the hyphenated term

"attitude-behavior, and related directly to Guttman's
definition of attitude as being "a delimited totality of
behavior with respect to something." According to Jordan
(l971a), this definition of Guttman's "is consonant with a
structural (Foa, 1966, 1968; Foa and Turner, 1970) approach
to the facet analysis of attitude-behavior" (Jordan, 1971a,
p. 7).

The facet theory approach fits within the positiv—

istic definition developed by McGuire (1969, p. 45)

and facilitates a cognitive-affective-conative

(knowing, feeling, and acting)ana1ysis of the
human condition. In the paradigms developed in

85

Tables 3-6, (i.e., Tables l6~l9 in Chapter V),
Level 1 and 2 deal with the cognitive component,
Level 3 with the affective component (evaluation),
Levels 4 and 5 with a combination of affective and
conative components, and Level 6 with the conative
aspect. In Rokeach's (1968) belief-value-attitude
analysis it could be argued that Level 1 and 2 deal
with beliefs, Level 3 with values, and Levels 4-6
with attitudes. Such a system of attitude-
behavior levels facilitates an analysis of the
interrelationships of the cognitive-affective-
conative trilogy as well as highlighting the
usefulness of the conative aspect (i.e., Level 6)
as the criterion measure of attitude (Jordan, 1971a,
p. 7).
McGuire (1969) discussed the extensive use of the three
divisions of attitudes, the cognitive-affective-conative
analysis. This trichotomy arose out of the existential,
philoSOphical thinking that man can take basically three
stances to human life: knowing, feeling, and acting.
McGuire (1969) defined the cognitive (perceptual,
informational, or stereotypic) component of attitudes as
how an individual perceives the particular object of the
attitude. It is the conceptual connotation that the person
gives the object, the 'stereotype' he has of it. Scott
(1968) defined this component as "the 'richness' of the
ideational content, or the number of ideas the person has
about the object" (p. 207). An individual's liking or
disliking of an attitude object is considered the affec—
tive (feeling or emotional) component of attitude. This
measure, an individual's affect-laden or evaluative content

toward the attitude object, is labelled by Scott as

affective salience. An individual's behavioral, motor,

86

or action tendencies toward the object of the attitude is
the conative (action, behavioral) component of attitude.
Scott called this dimension the overtness of the attitude
and states that it "implies the overt enactment of an
attitude is at least partly attributable to characteris-
tics of the person" (p. 207).

McGuire felt that the conative component may be the
crux of attitudes, since it is intended to measure actual
behavior of an individual. A problem arises in measuring
this dimension. Like the other two components, the method
of analysis usually chosen is the "paper—and-pencil
inventory which indicates how the person says he would behave
in the presence of the object, rather than by observation of
how he actually behaves" (p. 156). Past research indicates
there is often a low correlation between what a person
says he does and what he actually does. McGuire suggested
that further research might attempt to get at actual behavior
of the subject rather than self-reported behavior, to
improve the validity of the measurement of this highly
important component.

"An attitude, according to Allport's 1935 definition,
had at least five aspects: (a) it is a mental and neural
state, (b) of readiness to respond, (c) organized, (d)
through experience, (e) exerting a directive and/or dynamic
influence on behavior" (McGuire, 1969, p. 142). This defini-

tion encompasses the three components of attitude

87

(cognitive-affective-conative) postulated by McGuire (1969)
and applied by Jordan (1971a).

Elaborations and variants on psychometric theory
stem from factor analysis (see Cartwirght, 1964;
Cattell, 1961) and from the recent concern with
"moderator variables" (Ghiselli, 1956, 1960;
Saunders, 1956). Alternative models of measure-
ment have been proposed, such as Guttman's (1944,
1950) model of ordinal scaling, Guttman's (1953,
1960) image theory, Lazarsfeld's (1950, 1960) latent
structure model, and Coombs's (1953, 1964) theory
of data. But aside from Guttman's ordinal scaling
procedures, none of these models has found wide
application to attitude assessment. This may be
due, in part, to the belief that the rather
laborious procedures for data collection and
analysis associated with them are unlikely, under
present circumstances, to yield commensurate gains
in precision and validity. If this belief is
justified it does not necessarily inpugn the models.
Quite possibly the fault lies in our imprecise con-
ception of attitude or in the laxness of our require~
ments for validity. If these were more stringent,
it is possible that alternative models could be
demonstrated superior to classical psychometric
theory (Scott, 1968, p. 209).

Scott (1968) stated that attitude is a hypothetical
construct and because of this, cannot be measured directly.
It must be inferred from the subjects' responses. In
deriving a theoretical definition of attitude, Scott (1968,
p. 251), cited seven measures that he regarded as important
and crucial dimensions:

1. It would reflect the intended property veridically.

2. It would be unaffected by irrelevant characteris-
tics either within the subject or within the
testing situation.

3. It would not modify the property in the course of
measuring it.

4. It would make sufficiently fine distinctions
among persons to represent gradations along the
dimension as conceived.

88

5. It would yield results substantially equivalent to
those produced by another adequate instrument
measuring the same property.

6. It would yield equivalent scores on a retest
administered within a time period in which the
prOperty can be assumed to remain constant.

7. It would be relatively easy to construct, administer,
score, and interpret.

"Ideally, one would like to have precise, quantitative
indices for each of these criteria, in order to compare
available instruments and to guide the construction of im-
proved instruments" (Scott, 1968, p. 251).

The construct attitude has become so complex that one

can no longer talk clearly about "measuring an

attitude". Rather, one must restrict discussion

to procedures for measuring a particular property

of an attitude as conceptually defined. Whether

this degree of theoretical complexity in the

construct should be tolerated is a matter of some

concern. (hmamight hope that future conceptualiza-

tions will distinguish different constructs for many

of the properties now subsumed under the single term

(Scott, 1968, p. 265).

Kothandapani (1971) investigated the cognitive-
affective-conative trichotomy toward the prediction of
contraceptive behavior. In her review of previous findings,
she found few studies expressing a significant relationship
between attitude and behavior. Three reasons were postu-
lated by Kothandapani for this omission. Many of the
studies defined attitude simply as affect. Although
Kothandapani agreed that affect does play a large part in
attitude, she felt that it was not the sole component. A
second reason given for her belief in the lack of an associ-

ation between attitude and behavior was that often the

researcher is measuring attitude toward an inapprOpriate

89

stimulus. Thirdly, the instruments used to measure attitude
were often inappropriate or defective.

In her present study, Kothandapani (1971) has attemp-
ted to discover additional evidence for the converging and
discriminating entities of what she viewed as the three
components of attitude: (a) feeling, (b) belief, and (c)
intention-to-act. The first two, feeling and belief, were
the cognitive and affective elements. The behavioral com-
ponent was defined as 'intention-to-act' or 'behavioral
intention.‘ Her definition took into account behavior in
hypothetical situations; Kothandapani believed that this
wouldkxgmore powerful measure to predict behavior.

Intention-to-act was considered the most predictive
of the three elements of behavior, since 'intention-to-
act' and 'behavior' seem to share the most common deter-
minants. Kothandapani (1971) reported: "the hypothesized
convergent and discriminant validities of the tripartite
classification of attitude into feeling, belief, and
intention-to-act components were confirmed in this study"
(p. 331). The intention-to-act measure proved to be a
better predictor of behavior than either of the two other
measures. She felt that this was the reason most other
studies, which dealt primarily with cognitive items and not
measures of intention-to-act, failed to predict behavior.

Kothandapani (1971) viewed the tripartite distinc-

tion from the point of View of learning theory. The feeling

9O

component was seen as the conditioned stimulus, while
belief and intention-to-act constituted the discriminant
stimuli. Thus, the tripartite division of attitude could
be placed in the stimulus-response language of behaviors.
"Performance of an act may be considered a function of
learned intentions, beliefs, and feelings in combination
with current stimulus conditions" (p. 332). In behavior-
istic terms, the feeling element was seen as stimulating or
inhibiting performance, the belief division in forming
stimulus-response-reinforcer relationships, and the intention-
to-act element as an organizer of the feeling and belief
input to the behavioral output. Kothandapani saw this func-
tional approach to the feeling-belief-intention-to-act
system as accounting for the superior predictive power that
she found for the intention-to-act component.

Another representation of the tripartite classifi-
cation of attitude, iS illustrated on page 91, where
attitudes are broken down into measurable independent
variables, intervening variables, and measurable dependent
variables (Rosenberg and Hovland, 1960, p. 3).

Guttman, in 1959, proposed a structural theory for
inter-group behavior. This was perhaps the first step in
the union of structural theory and applied behavior. In
regard to structural theory, Guttman discussed some of the
reasons he felt that empirical correlations do not form

perfect relationships with one another. He stated that

 

H0H>mcwh mchHmocoo

 

 

 

 

 

 

91

 

 

 

mucmEmumum Hmhum> uoH>mcwm
mGOHuom uum>o
. A=muomflno
mmmHHmn ,
coa ammo .1!) - mcsuHuum= Hmsuo cam
mo mucmEmumum Hmhum> .u. o 4 . (I. mmcsuHuumuill .mmsoum Hmwuom .mmsmmH

 

 

 

 

 

mmmcommmu Hmoumooumm \\\\

 

HMHOOm s mGOHUMSHflm

 

 

powwow - .mamnee>eech Hesseum

 

 

 

 

mo mucmEmumum Hohnm> om ,
noncommmn u mmd

mso>Hmc OHumcummawm [I‘ll/i

 

 

 

 

 

mmHQMHHm> mehmHHm> meannm>
ucmccmmmc mHnmnsmmmz mchm>nwucH usmccmmmccH anmudmmmz

mmDDBHefifl m0 ZOHBQMUZOU UHdemmUm

92

there are many subvarieties of behavior and that these sub-
varieties had been lumped together in most applied attitude
research, producing an intermingling of categories which
should have been considered distinct entities from each
other.

It is also the task of the social scientist to
discover the structure underlying the totality of behavior.
The relationships between subuniverses of behavior must be
discovered, so that each separate behavior class may exist
independent of the others. The total structural system,
including all these subcategories of behavior will be called
the universe, each subcategory of behavior by itself repre-
senting a subuniverse of subvariety of the universe of
behavior. Certain subcategories of behavior will be more
closely related to one another than others and some less
related. "A task of the social theorist is to provide an
abstract framework whereby to define the subuniverses: the
more adequately it explicates the empirical correlations
that ensue among the definitions, the better the framework"
(Guttman, 1959, p. 318). Guttman's three-facet, four-Level
system was an attempt to provide this basic framework.
Improving the empirical correlations among the Levels is
what Jordan has attempted to do, by adding two Levels and
two facets and refining the items in the scales he has
developed. "Comprehension of the multivariate system of

the universe can lead to larger theories of relations with

93

other universes, and thus to more and more perfect multiple
correlations for each variety of behavior separately"
(Guttman, 1959, p. 318). Jordan has been attempting this

with his Attitude Behavior Scales. The first scale involved

 

the mentally retarded as the attitude object.

Summary
Guttman (1959) noted that "proceeding from a

semantic structure to a statistical structure appears
necessary in order to relate abstract social theory to
empirical research" (p. 319). This was what Guttman attemp-
ted; his conceptualizations were based on strict, methodo-
logical rigor founded on developments following the factor
analytic techniques of Spearman and Thurstone. Through
semantic restructuring of the variables in Bastide and
van den Berghe's (1957) research, he rearranged the design
and applied statistical analysis to the data. Bastide and
van den Berghe's work was consistent with the cognitive-
affective-conative analysis dating back to early Greek
philosophy. Jordan (1968) found the system lacking
especially in the 'conative dimension' and somewhat in the
'affective domain;' thus, he extended the four levels to
include a more comprehensive view of the cognitive-affective-
conative trichotomy.

Guttman's "continguity hypothesis" prOposed that

variables of Levels with more similar facet elements are

94

more related than those with less similar elements. Levels
closely related to one another may form a hierarchical
relationship or simplex. The interrelationships among
variables will be posited by the hierarchical order among
the facets (Foa, 1968).

Guttman (1967) proposed the following faceted
definition of varieties of racism:

((belief) ((1ndividual's)

"Racism = that the

z(action) 2(individual's group)

(comparative)

behavior of a nature vis-a-vis another

(interactive)§

group it regards as racially different from itself

(ought to be)l

unfavorable to the group presumed to be

(actually is))
racially different."

"To increase the predictablity would require
enriching the facet design (i.e., of the previous varieties
of racism), or placing these behaviors in a larger context"
(Guttman, 1959, p. 327). Chapters III and IV explore
racism, prejudice, and color-labelling as prerequisite to

developing a more complete facet design, in Chapter V,

for the study of racial attitude-behaviors.

CHAPTER III

RACIAL ASPECTS OF PREJUDICE

Racism is not a new phenomenon. Men have long tried
to identify themselves not only as individuals but

as members of social groups; and to set up viable
social groups, they have thrust others out. These
'others' have been differentiated in various ways,
for instance, according to clan, tribe, nation, estate,
or class. The forms change, but the process of self-
definition is seemingly endless. And all these
'others' have one feature in common: they are never
quite as good as the self. Some mysterious tag of
devaluation is attached to the other person as his
essential point of distinction from the groups of
selves (Kovel, 1970, pp. 13-14).

Introduction

 

This chapter discusses how racial aspects of prej-
udice relate to race labelling, language, and the cognitive,
affective, and behavioral aspects of attitude. There is a
definite interaction between race labelling and prejudice;
research indicates that both contribute to the negative
images white people have for blacks. The exact relationship
between racial labelling and prejudice is still vague. To
understand race prejudice, recent theories have attempted
to analyze attitude through division into its three compo-
nent levels: cognitive, affective, and behavioral. The
historical origins of this tripartite classification were

discussed in Chapter II. Jordan's (1971a) research is one

95

qt—e

96

of the sole theories to posit a tripartite attitude theory
that also instruments a statistical analysis of these three
components. The present study is an attempt to relate
Jordan's theory to research begun by Williams (1964, 1966,

1969) on racial concepts and language.

Race Labelling»

 

One major outcome involved in the process of 'race
labelling' is the process of creating a stereotype. As will
be evident in the following discussion, not all 'race
labelling' creates stereotypes, but a large proportion of it

"a

does. Vinacke (1956, p. 107) defined stereotyping as
tendency to attribute generalized and simplified character-

istics to a group of people in the form of a verbal label."

(The black man in the United States has been given a label as

'black,‘ 'dark,‘ 'colored,‘ 'Negro' at various times in the
history of this country. (With many of these labels has come
negative connotations, negative stereotypes that have
accompanied these labelsilgThese labels carried with them
society's attempts to devalue the ability, power, and pres-
tige of people so labelled.) As one observer noted, blacks
"discovered that their determination to do the right thing
had blinded them to the wrgng thing 'the man' was doing, as
he suggested to them how important it was to be a good 'Negro'
teacher or a very fine 'Negro' social worker or psychologist

or psychiatrist." In large part, black people failed to see

4’
“I

.x

97

the true meaning given the adjective 'Negro' and the
decreased value that had been placed upon it (Grier and
Cobbs, 1971, p. 129).

M. B. Smith (cited in Baughman, 1971) felt thatithe

word and category 'Negro' is an American creation, closely
associated with the slave period. The term does not des-
cribe nor fit blacks in Africa, and he suggests that black
Americans should refrain from using the term, and that

white Americans do likewise. The term 'Negro' was imposed

\ , r.“

upon black individuals; whereasxthe label 'black' was a self-

c.

, ' ‘\
chosen one, one that connotes respect and pride.)
’3 f

n 4 In a study conducted in 1968 on the South Side of

Chicago, blacks preferred the term 'Negro' to other options,

-.\

)
such as 'black,’ 'Afro-American,‘ and 'colored'](Baughman,
1971). However, Baughman felt thatithe younger, more

articulate, and more militant blacks prefer to be called

\\ n g"

'black' rather than 'Negrq;' the term 'Negro' is often

seen as a label imposed upon blacks by the white society,
\
instead of a voluntarily chosen one.3

Baird (cited in Bennett, 1967) objected to the use
to

of the word<?Negro' because of its slave orientation and
involuntary imposition and its lack of geographic or cul-
tural specificity. )Moore (cited in Bennett) agreed that

blacks should not be referred to as Negroes; he felt

(f'Negro' was odious and filled with negative stereotypes.

\

‘ ('Black,' he felt, was a more positive concept than 'Negro,'

98
\

but was still associated with negative connotations::>Closer
identification with African culture and heritage seemed to
be the next step that both Moore and Bennett would desirea;
The most appropriate term to use would seem to be the one
that is most voluntarily chosen and desired by blacks.

2; Q The word 'Negro' was not fully accepted as the desig-
nation for black peOple until early in the Twentieth Century.
Acceptance of it was very slow until that time, but after it
became accepted, there was little opposition to the term from
either blacks or whites. The NAACP, founded in 1909, chose
the term 'colored' over 'Negro')(§222y Photo-editorial, 1968).
Recently, pressure has been put on the NAACP by black power

advocates to change its name to NAABP~~Nationa1 Association

for the Advancement of Black People.

“A,

I
¢

(\A survey was conducted by Ebgny magazine during 1967-
1968, to examine the choice of a term by subscribers of the
magazine. Although the authors stated that the study in

no way involved scientific sampling, the results were inter-
esting. Of the 2000 people who responded (one-tenth of one
per cent of EEEEX readers), 48 percent preferred the term
'Afro-American;' 23.3 percent preferred 'black;' 12 percent
preferred 'African American;' 8.1 percent preferred 'Negro'
and 3 percent preferred thechoice 'colored' (Ebgny.Photo-
editorial, 1968, p. 164):€:This survey indicated that blacks

definitely do not prefer the term 'Negro' any longer.2>

99

Jordan (1971a) and Hamersma (1969) felt that presently
the black community prefers use of the word 'black' over

'Negro.' They constructed Attitude Behavior Scales for

 

measuring racial attitudes of blacks and whites toward each
other. The scale given to blacks, measuring their attitudes
toward whites, uses the word 'black,‘ while the scale given
to whites and measuring their attitudes toward blacks,
contains the word 'Negro' instead of 'black' as the

attitude object. Jordan and Hamersma\(1969) explained that
this procedure was followed because of<general acceptance

of the word 'black' in the black community, while the white
community still seemed more familiar with the term 'Negro.'}
At least for college students, it would appear that the
term 'black' can now be used in testing white populations;
the black power movement has aided in familiarizing white
college students with its use.

Williams and Kirkland (1971) differentiated between
the words 'black' and 'Negro' (see Chapter I, Introduction,
pp. 6-7). Further comment on the 'black' versus 'Negro'
versus 'Afro-American' controversy follows the Williams and
Kirkland discussion (through p. 9).

Dubois (1961) commented on the internal struggle of
black identification: "one even feels his twoness~~an
American, a Negro; two souls, two thoughts, two unreconciled
strivings; two warring ideals in one dark body whose dogged

strength alone keeps it from being torn asunder."

100

./

According to Lester (1968, p. 92); "blacks now
realize that 'Negro' is an American invention which shut
them off from those of the same color in Africa. They

recognize now that part of themselves are in Africa?) This

I
1

new identity is an attempt to replace and recreate ideals
that had been lost to them . . . they have stOpped being
Negroes and have become black men, in recognition of this
new identity" (Lester, 1968, pp. 91-92).

/L JReview of literature indicated that although there

was disagreement as to what term to use, most studies

agreed that the term 'Negro' was no longer an appropriate
W!

J

label to use.}iThe terms presently preferred were 'Afro-
American' and 'black." The associations of 'Negro' with
slavery were evident; voluntariness of choice is important
in the choice of the most appropriate label for a peOple.é
As Campbell (1971, p. 1) has found, "there is no 2
simple way to describe white attitudes toward black people.
In the past, when slavery was present, it was possible and
probably predominant for white peOple to have a common,
stereotypical image of blacks. With the emergence of the
Black Power Movement, encompassing the concept of black
pride and capability, it is no longer possible for whites to
look upon blacks as all similar. From daily observation,
this conclusion by Campbell must be questioned. Many whites

still see blacks in a stereotypical manner. This may be

less true in the cities, but in rural areas where contact

101

with blacks is minimal, whites quite often stereotype
blacks and their behavior.

In an analysis of attitudes of whites toward
blacks, Campbell (1971) found a great variety of attitudes
concerning race. About 25 percent of the peOple surveyed
were generally positive in their attitudes about blacks,

25 percent were generally negative, and the rest were some-
what between the two polarized groups. The most positive
change in attitude they found was in those individuals who
had attended or were attending college. College students'
(those individuals who had attended college within the past
twenty years) views had become quite different from their
parents' views. Campbell speculated that this change was
due, in part, to a liberalization of views brought about
with the New Deal politics, the growing importance and
interest in the social sciences, and the social changes
brought about by the Depression.

In their report, the Kerner Commission stated the
belief that this country is moving toward two separate
societies, one white and one black. Campbell (1971) felt
that this was not necessarily true; he found many instances
where white and black were moving closer together. In the
reports from the Institute for Social Research at the
University of Michigan, no dramatic changes were found in
feelings of the two racial groups toward each other.

Rather, they found an increase in racial contacts, more

102

persons of both races indicating that they had friends of
the other race. Although there was somermwnhugapart, they
concluded that, in general, whites and blacks had not
begun to establish two split societies or philoSOphies.
One thing to be emphasized, however, is that there were
great differences and prejudices between blacks and whites
that both races must attempt to solve.

The most consistent changes Campbell (1971) found
were in those individuals who had attended college some-
time after the Second World War. Those who had not gone
past a high school education appeared to have the most
resistant attitudes. In the area of education, Campbell
concluded that a college experience was crucial to modifying
past negative white attitudes to a more positive stance.
This indicated that not enough emphasis was being placed in
the educational institutions before college to lessen
prejudice. More effort could and should be made in the
grade schools and high schools to bring the races closer
together.

In a discussion of the future of racial attitudes,
Campbell (1971), stated that the most dramatic shift in
attitudes of white Americans has occurred in the period
after World War II. Conclusions of the three national
surveys by the Institute for Social Research, found that
interracial contact was definitely increasing. They also

discovered that there has been a greater proportion of

103

blacks employed in white-collar jobs and there was increased
educational achievement among young blacks. The survey team
concluded that a college education leads an individual
toward more liberal views on racial concerns, the future of
white attitudes toward black people cannot be considered
without reference to black attitudes toward whites--
each influences the other (Campbell, pp. 159-162). "The
collegesenxapouring out successive cohorts of young people
whose racial attitudes are in the large clearly more
positive than those of the pOpulation into which they are
moving" (Campbell, p. 160). They concluded that these
attitudes will not revert back to their previous stance, and
therefore, these changed views will have a definite effect
in changing prevailing American attitudes.

Campbell (1971) felt that the black population
will assimilate into the mainstream of American life, as
other "minorities have done," but predictedtflmnzthis assim-
ilation will be a difficult process and will not be achieved
in the present decade. "The white population is far from
a general acceptance of the principle and practice of racial
equality . . . we are at present at a point of uneasy con-
frontation. American society is developing a new pattern of
relationships between white and black and the period of
change is a time of tension for both races" (p. 162).

King (1970) noted the need for white society to

recognize racism in its midst and understand its presence

104

before it can move to eradicate it. Blacks are attempting
to change it, but are finding much resistance from the

white populace. "The white majority, unprepared and unwilling
to accept radical structural change, is resisting and pro-
ducing chaos while complaining that if there were no chaos,
orderly change would come" (p. 8); He saw the period from
1955-1965, as a time, although somewhat constructive, that
mislead both blacks and whites in this country. "Everyone,
activists and social scientists, underestimated the amount
of violence and rage Negroes were suppressing and the amount
of bigotry the white majority was disguising" (p. 9).

King summed up the recent racial disturbances and riots

with a quote from Victor Hugo written a century ago: "If

a soul is left in darkness, sins will be committed. The
guilty one is not he who commits the sin, but he who causes
the darkness". "The policy makers of the white society

have caused the darkness; they create discrimination; they
structured slums; and they perpetuate unemployment, ignor-
ance, and poverty" (p. 9). Although he viewed inner-city
crime as wrong, King saw it more as derivative crime; crime
derived from racist and suppressionistic policies of the
white society. All the laws violated by whites in the areas
of welfare, building codes and regulations, police tactics,
and education, led to counter crimes by black inner-city

residents.

105

Goldschmid (1970) found that white prejudice toward
blacks was much greater than black prejudice toward whites.
One particular study, Noel and Pickney (1964, cited in
Goldschmid), found black prejudice toward whites signifi-
cantly less present than white prejudice toward blacks:

41 percent of blacks in the study received the lowest
possible prejudice score, compared to only 5 percent of
whites; 17 percent of blacks received the highest possible
score, compared to 48 percent of the white subjects tested.
These results held across all socioeconomic levels.
Strickland (1965, cited in Goldschmid), in an analysis of
prejudice in college students, found less social distance
toward the other racial group, in black students than in
white students. Goldschmid cited three additional studies
that found similar results (Bogardus, 1958; Bryant,
Gardner, and Goldman, 1966; and Webster, 1961). He concluded
that "it is clear that black prejudice towards whites,
although much more easily justifiable in View of centuries
of white oppression of blacks, is much less common than
white prejudice toward blacks" (p. 257). A recent survey,
however, noted that although white attitudes have become
more positive (between 1956-1971), black attitudes have
become more negative in a particular midwestern city

(Detroit Free Press, 1972). This may indicate a change in

 

racial attitudes; black attitudes may have become more

explicit in recent years. Blacks seem more willing now to

106

express their honest opinion regarding racial attitudes;
also, recent changes in American society have made it possi-
ble for members of a minority to express their views without
fear of reprisals.

A study completed by the United States Commission of
Civil Rights (1970) stated the important relationship between the
racial composition of schools,and attitudes. Whites attending
integrated schools were more interested than whites attending
ckafactOsegregated schools in seeing passage of equal oppor-
tunity laws, expressed more willingness to reside in an
integrated neighborhood, and eXpressed more willingness to
make friends with black people. More legislation could be
enacted to increase integration of schools; the early years
of school for children are ones in which attitudes can be
most greatly affected.

Pettigrew (1964) examined what the concept, 'freedom'
means to black Americans. Pettigrew saw freedom as a
cessation of adopting the inferior, subordinate role, a
casting off of all the remnants of stereotyped inferiority.
He saw it as equally important for black peOple to realize
that they have been playing this inferior role, as it was
for whites to realize that they have been enforcing this
subordinate position on blacks. To see each other as
human beings with human needs and desires, both races

must break down the facade contained in maintaining role

barriers.

107

In a discussion of what Goldman (1970) referred to
as 'revolution,‘ he noted the positive aspects of overt
acting-out behavior. He saw riot behavior as a desperate,
last hope attempt to obtain freedom, whereas 'revolution' he
viewed as "an act of hope; and a revolution which prefers
negotiation to war, which seeks not to destroy but to
reform America and which continues in spite of everything
to operate on the premise that it is winning is the most
transcendently hopeful act of all" (Goldman, p. 203).

J. Boggs (cited in W. G. Smith, 1970, p. 144), very
succinctly noted the relationship between human material
wealth and human responsibility, and the values that have
been placed upon each by Americans:

Coming in the United States at this time, when

there is no longer any problem of material

scarcity, the Negro revolt is therefore not just

a narrow struggle over material necessities. It
does not belong to the period of struggle over goods
and for the development of the productive forces,
which we can call the era of Dialectical Materialism.
Rather, it ushers in the era of Dialectical Humanism,
when the burning question is how to create the kind
of human responsibility in the distribution of

material abundance so that everyone can enjoy and
create the values of humanity.

 

 

As Schwartz and Disch (1970, p. 4) put it, "the
condition of the black man in America reveals that whites
consider their pocketbooks to be far more important than
their morality."

Blauner (1970) rejected the idea that blacks will

or should be assimilated into the mainstream of American

108

culture. He felt that this is what most whites want to
believe; even social scientists hold to the concept of
assimilation. Blauner's Opinion was that the experience of
blacks in America had little in common with the experience
of other minority groups. The experiences of a minority
group entering the dominant society to a great extent
determines how the two cultures will react to one another.
For the black man, slavery destroyed their African group
identity. Even during this destructive process, blacks
began re—building a cultural and ethnic-group identity.
(Blauner (pp. 112-113) defined black culture as "not just

poverty, black ghetto, lower-class, but also "a complex
mixture, whose sources include also Africa, slavery, the
South, Emancipation and northern migration, and above all,
racism . . . Though this culture is overwhelmingly the
product of American experience, the first source is still
African." It was American racism that Blauner (1970)

felt led to the develOpment of a black culture within
America.>(The struggle against this racism provided the
basis from which black culture grew and flourished)
Blauner's arguments against black assimilation into the
mainstream Of American life are not totally justifiable.
Many minority groups have assimilated into American culture
while still maintaining their distinct sub-culture charac-

teristics. There is little reason to assume that blacks

could not make a similar transition without losing their

109

African ethnic-group identity. White acceptance of blacks
and black culture would help facilitate this change.

J. H. Clarke (1969, p. 16) poignantly portrayed the
dilemma the black man was placed in when he was transposed
from his African homeland and thrust into a country totally
foreign to his being and life—style.

L (”The Africans . . . were neither respected Africans

‘ nor accepted New World Americans. They were renamed,

and became a marginal branch of the human family
now referred to as NegroesJN The Europeans needed a
rationale for their actions*and a rationale was
created with supporting concepts. The cruelest
concept ever devised by the mind of man was created
to support the slave trade and the colonial system
that followed-~the concept of race and the assump-
tion that there are superior and inferior races.
The Africans were depicted as people without a
history who had never properly handled power and
who, certainly had made no contribution to the
development of human cultures. And thus, the
seeds of the present day conflict were planted.

In a discussion of black self-concept, Proshansky
and Newton (1968) stressed the dire effects of slavery and
the caste-class system that followed. They saw the conse-
quences of slavery as resulting in a double burden for black
Americans: one being the devastating psychological effect
and the second being the social and economic inequities.
Besides the social and economic impositions of poor and
inadequate housing and schools, unemployment and under-
employment, inferior jobs, was the less visible but equally

serious consequences Of low self-esteem, identity conflicts,

and feelings of helplessness and hOpelessness.

110

Seward (1956) insisted that blacks must be judged
from their own subculture, not from standards set up by the
dominant white culture. By viewing black individuals from
the dominant culture, deviations are seen that are not
deviations when living or considered from the values of
the subculture.

The Relationship Between Language
and Racial Prejudice

 

 

According to Podair (1970) language can have a
definite influence on the learning of prejudice. Language
can shape ideas and concepts in both conscious communication
and unconscious thought. It has contributed to the stereo-
types generated against the black individual and the develop-
ment of racial prejudice (Podair, 1970). The studies by
Osgood and by Williams and his associates have shown how
important the meanings of words are and how they can give
previously neutral concepts negative connotations. Podair
(1970, pp. 388-389) stated that concepts such as blackball,
blackbook, and blacklist, "cannot be considered accidental
and undoubtedly would not exist in a society wherein whites
were a minority. Historically, these concepts have evolved
as a result of the need of the dominant group to maintain
social and economic relationships on the basis of inequal-
ity if its hegemony was to survive."

Podair (1970) discussed the connection between the

negative affective meanings associated with the word black

111

as it effects the black child, and how the unfavorable
connotations effect as well the white child's perceptions
of the meaning of black. These negative meanings are then
carried into and through adulthood.
The term 'black sheep' is utilized in all segments
Of our society to denote an individual who seems
almost completely immobilized in his efforts to
effect an adjustment to his environment that could
result in fulfilling personal, familiar, and
community responsibilities . . . The acceptance
Of the term 'black sheep', however, assists in
the creation of a stereotype of Negro 'irrespon-
sibility' which has become so valuable to the
advocates of prejudice (Podair, 1970, p. 398).

Podair (1970) saw the language of a society as
reflecting its social values and attitudes. As a society
changes, the thinking and speech patterns will also change.
Thus, the symbolism behind the concept black and white may
change as our society rids itself of prejudice. "Until
that time, however, the relationship of language to preju-
dice towards the Negro will be of import to the social
scientist working to meet the challenges arising out of the
problems Of race relations in present day America" (Podair,
1970, p. 391).

Another view1 is the argument that the negative
connotations of the word 'black' may persist even in a
society free of racial prejudice. The negative connotations

might be seen in a Manichean sense which describes the

forces Of darkness and light in Opposition.

 

1Personal communication with Dr. Maryellen McSweeney,
College of Education, Michigan State University, Oct. 5, 1972.

112

Following this line of reasoning, it would be
conceivable that a people could be called 'black' without it
having negative connotations attached ot it. The color
'black' might still be used for things considered negative,
but this does not necessarily have to generalize to the
black peOple.

Citron (1969) discussed what he labelled, "the
whiteness of the world Of the white child." The white child
living in a world of whiteness builds feelings of the
"rightness Of whiteness" into his personality. Everything
associated with white and light colors (i.e., white skin)
is perceived as acceptable and good, while dark skin and
dark colors are associated with quite the Opposite. As
the child develops, this feeling of "white is right"
becomes what Citron calls "a white-centric world." The
norm is white; others Of different color are judged by it.
This develops into an inferior—superior dichotomy. K.
Clarke (1963, cited in Citron, p. 4), noted that ". . .
children's attitudes toward Negroes are determined chiefly
not by contact with Negroes but by contacts with the pre-
vailing attitudes toward Negroes. It is not the Negro
child, that influences children." Citron cited studies by
Goodman indicating that white children, early in life,
develop an emotional rejection of blacks. One positive note,
however, was that a recent study (Triandis, Malpass, and

Davidson, 1972) found that black children no longer chose

113

the white doll, but indicated a preference for the black
doll in a laboratory situation; earlier studies by Clark
(1947) indicated that black children chose the white doll
over one of their own color.

As Williams' (1964, 1966, 1969) studies indicated,
white is seen as a symbol of rightness, cleanliness, goodli-
ness, purity and beauty, while black is seen as sin, dirt,
and impurity. Historically, white has dominated cultures;
the children expressed the same racist notions imbedded
within their particular culture (Citron, 1969). Citron
noted that the fear and rejection of black becomes established
in children long before they have the 'rational' content
to support these ideas.

Centuries of white imperialism over darker peoples,
over three hundred years of the institution of
slavery in this country, and a quasi-caste system
since the days of Reconstruction, have produced
concepts and language forms fitting the needs of
the dominant group. These forms play their part
in forming the habits of thought of children.

There has been generated a mythology of racism,
with its stereotypes of primitiveness, immorality
and dangerousness (Citron, 1969, pp. 13-14).

Citron (1969) felt that one language form used to
encourage the racist philoSOphy was the "contrast-terms,"
black and white. These Objective words are given super-
ordinate, subordinate meaning and emotional affect. This
was similar to the general conclusions Williams and his

associates had reported (cited in Chapter IV), except that

Williams attributed more to the actual negative value of

114

the color, 'black.‘ Thus, it can be seen how crucial words
and language can be used in the perpetuation Of a racist
system. As Citron saw it "racism invests skin color with an
enormous and completely irrational salience in our country"
(p. 14). Although the literature shows how the color 'black'
has been given negative connotations throughout history,
it appears that prejudice and racism in this country has
given black people a negative connotation, not just the
color, 'black.‘
As Citron (1969, p. 17) concluded:

It is Whites as a group who enforce the repressions

Of the racist system and every White, especially

those in middle and upper class positions, because

they have more political and economic power, should

be actively involved in destroying racist arrange-

ments, practices, exclusions, double standards, folk-

ways and institutions, and should be actively involved
in building the conditions of equality.

Cognitive, Affective, and Behavioral Aspects
of Racial Attitudes

Mann (1959) in a discussion of the relationship among
cognitive, affective, and behavioral aspects Of race prej-
udice, stated that few studies have attempted to consider
all three Of these aspects and whether or not they are
related to each other. His study found possible evidence Of
a relationship among the three, but he stated that the
evidence for the relationship was not conclusive. For
black college students, he found a positive relationship
among the three aspects of racial prejudice. However, for

white students no positive relationships were found and a

115

negative relationship was found between the affective and
behavioral aspects Of race prejudice.

Insko and SchOpler (1967), in a paper on triadic
consistency reported the philOSOphy of the relationship
among attitudes, cognitions, and behaviors. According to
this theory, there was a tendency for these three aspects
to be related. The authors stated that, in Operational
terms, "there is a probabilistic relation between holding
certain beliefs and attitudes and manifesting certain
behaviors" (p. 366). Kothandapani (1971) also cited
evidence indicating a relationship among these three com-
ponents of behavior. In a schematic View of attitude,
Rosenberg and Hovland (1960), cited affect, cognition, and
behavior as the three measurable dependent variables of
the intervening variable of attitude.

Jordan's (1971a) work has attempted.to show the
relationship between cognitive, affective, and behavioral
components of attitude and has developed a simplex struc-
tural relationship among these aspects in a number of
studies involving race and attitude (Frechette, 1970;

Jordan, 1971a, 1972; Williams, 1970).

Conclusion

 

Racial differences are comparatively minor, except
for a few physical features. It has been man himself who

has blown these differences entirely out of proportion

116

and hence, created artifically marked distinctions among
the races. Through labelling people as different, man
began the process that led to what we now refer to as
racial prejudice. A race is given a label, the label is
then associated with negative connotations, and inferior
and superior races are created. The actual labels them-
selves may mean very little; it is the emotional concep-
tions underlying them that give them their power. It is
not so much the labels that are used, but how they were
implemented and what they mean-~the attitudes behind the
labels.

Perceived voluntariness Of choosing a particular
racial name is one important variable as to the meaning
invested in the term. The ability to create pride in the
label is another important feature of racial terminology.
Both Of these variables are important for the race so
named to develop and maintain positive attitudes of self.
Once a race has a positive self-concept, it becomes more
difficult for other races to devaluate them. Thus, although
a racial label may play a small part in the formation of
attitudes and prejudice, its presence is indeed felt.

Angeles (1971) in his book on understanding the
black eXperience, perhaps summed up best the dilemma of
the racial situation between black and white peOple in this
country, a dilemma he felt should not exist. He introduced

his book thusly: "This book is absurd. Its author assumed

117

that it can help non-blacks begin to realize a love for the
black, but this love they should already have, if for no
other reason than that we are all alive and struggling for
a life Of decency and brotherhood" (p. 11).
This is but one area of research concerned with

attitudes. Before we can reach any definite conclusions

we must arrive at a more refined definition of attitude.
The Guttman-Jordan system is an attempt to do this. Their
analysis includes not only theoretical notions of attitude,
but also a semantic methodological system of measurement

accompanying their theoretical conceptions.

CHAPTER IV

RACE, COLOR-LABELLING, AND THE

SEMANTIC DIFFERENTIAL

Black is when they say '. . . one nation indivisible
with liberty and justice for all . . .' and you
wonder what nation they're talking about (T. Brown,
Jr., 1969).

Introduction

 

The history of word symbolism illustrates how 'black'
has been associated with negative connotations, while 'white'
has usually been associated with positive characteristics.
Osgood (1957) developed the Semantic Differential to measure
word meanings. Williams (1964, 1966, 1969) applied this
Semantic Differential technique to the study of color sym-
bolism and race. His research has supported the observa-
tion that, in American culture, black is generally associated
with 'badness' while white symbolizes 'goodness.;> Other
studies (Jenkins, Russell, & Suci, 1958; Osgood, 1973;
Williams, Best, Wood, & Filler, 1971; Williams, Tucker, &
Dunham, 1971; Lessing & Zagorin, 1972) concur with this
conclusion. Williams and his associates tested these
results in the laboratory (Williams & Edwards, 1969) and

also cross-culturally (Morland & Williams, 1969; Williams &

118

119

Carter, 1967). Two of Williams' recent studies (Williams,
Best, Wood, & Filler, 1971; Williams, Tucker, & Dunham,
1971) evaluated the changes occurring in word connotations
between the years 1963 and 1969. These studies related some
interesting findings that ran contrary to some of Williams'
earlier results. Osgood (1973) discussed universal trends
in color preferences, connotation, and association. This
study supported the notion that blackness and 'badness' are
associated, and white and 'goodness' are associated, and

that these associations still exist and exist cross-culturally.

Word Symbolism

 

Racial overtones associated with the words and con-
cepts used for blacks and other non-whites were evident in
the language of the very earliest settlers of America.

This special language was used first to define lower status
for non-whites and later to justify the status that had

been assigned to them (Schwartz and Disch, 1970). In an
historical anthology and extensive review of the literature
on the roots of racism, Schwartz and Disch (p. 6) stated that
"before the close of the fifteenth century, the words
'soiled' and 'dirty' first became the linked with word
'black'. By 1536, 'black' connoted 'dark purposes',
'malignant', and 'deadly'; by 1581, 'foul', 'iniquitous';

by 1583, 'baneful', 'disastrous', and 'sinister'.' Osgood's
(1973) cross-cultural study indicated that black and dark

colors universally represent negative perceptions.

120

Even the Bible contains associations Of black and
badness, white and goodness. Throughout, similarities can
be found linking words such as black, evil, damnation,
despair, and sin. The New Testament has similar associa-
tions between blackness, darkness, and evil (Schwartz and
Disch, 1970). "The impact of biblical color imagery was
inevitably reflected in the works of the great English
writers from Chaucer to Milton" (Schwartz and Disch, p. 7).

The symbolism involved in the two terms has differ-
entiated black and white, so that these words have become
polar opposites. As Fanon (1967) has stated, whiteness
has been associated with all things good (i.e., white dove
of peace, the bright look of innocence) while blackness
and darkness have been associated with quite the opposite
(i.e., abysmal depths, the labyrinths of the earth, blacken
one's reputation). Fanon continued to state that in all
civilized countries, the Negro is the symbol of sin; a
white child is looked upon with much admiration and joy,
while the black child receives none of this adulation and
respect. He did not define what he meant by 'civilized'
countries, but it appears he meant 'industrialized' as
Opposed to 'developing' (or 'pre-industrialized') nations.

R. L. Williams (1972), in a discussion of the
changing image of the black American, stated his belief in
the importance of terminology in the images that are asso-

ciated with blacks. By modifying a noun with the adjective

121

'Negro,' it sets it aside from the dominant group. "The
adjective does what it is supposed to do, 'describe and
limit.‘ The prefix 'Negro' establishes a great deal Of
restrictions on the noun it modifies" (p. 68). Discrimina-
tion in the areas Of jobs and housing in America supports

these conclusions of Williams.

The Semantic Differential Technique

 

The most extensive studies using the Semantic
Differential technique as a measure of semantic meanings
was undertaken bytflmzoriginator Of the scale, C. E. Osgood,
and his associates at the University of Illinois. J.
Williams and his associates, using the Semantic Differential
techniques, applied the scale technique to the concepts of
race and color.

The rest of this chapter will be devoted to a
discussion of the Semantic Differential technique as devel-
Oped by Osgood (1957) and the applications devised by
Williams (1964, 1966, 1969) and subsequently by Lessing
and Zagorin (1972)-

Osgood, Suci, and Tannenbaum (1957) discussed the
methods through which the Semantic Differential technique
was devised and the meaning and reasoning underlying it.
They saw the concept Of 'meaning' as a relational device
and have attempted to establish psychological meaning for

concepts which are then applied to use in the Semantic

122

Differential. As they saw it, "the Semantic Differential
relates to the functioning of representational processes
in language behavior and hence may serve as an index of
these processes" (Osgood, Suci, and Tannenbaum, 1957, p. 9).
They saw language signs as having certain psychological
meanings and with this prOperty can be used in a consistent
manner in situations to produce certain behaviors that will
then contain social meaning. The consistency of behavior
in situations also gives meaning to the processes that
are represented semantically.
The Semantic Differential is an attempt to measure
a certain type of meaning (attitude) through a small sample
of words that vary only along the dimension being measured
and are largely insensitive to any other sources of variation.
A concept in the semantic Space is defined by mediating
variables composed Of antagonistic pairs and varying along
an intensity variable. Each polar word group in the semantic
scales is associated with mediating processes, depending
largely on the polarity of the terms, and its intensity
determined by the seven categories (spaces) in between
the two words of the pair. Which of the seven spaces
chosen by a particular individual depends upon a media-
tional process of selection involving an intensity variable:
Through the functioning of a generalization principle,
the concept will elicit checking of that scale posi-
tion whose dominant mediator component most clearly

matches in intensity the corresponding component
in the process associated with the concept itself.

123

Since the positions checked on the scales consti-

tute the coordinates Of the concepts' location in

semantic space, we assume that the coordinates in

the measurement space are functionally equivalent

with the components Of the representational media-
tion process associated with this concept (Osgood,
Suci, and Tannenbaum, 1957, p. 30).

NO general Semantic Differential scale has been
devised; rather, each specific research project using the
scale should adopt Semantic Differential words which are
both representative and relevant to their particular area
of interest. Words should be chosen that take into account
individual differences, representativeness, relevancy, and
semantic stability (Osgood, Suci, and Tannenbaum, 1957).

Reliability for a Semantic Differential high on the
evaluative factor, was established through test-retest
correlation data. For one hundred subjects who took forty-
item scales, the reliability coefficient was equivalent to
.85 (Osgood, Suci, and Tannenbaum, 1957).

Osgood, Suci, and Tannanbaum (1957) stated that
since there was "no commonly accepted quantitative criterion

Of meaning,‘ 'face validity' must be used as the sole
measure of validity for the Semantic Differential. Validity
studies have been run on the 'evaluation' dimension. Two
comparison studies were reported by Osgood, Suci, and
Tannenbaum, one comparing it with Thurstone scales, the
second with a Guttman—type scale. "The correlation between

the Semantic Differential scores and the corresponding

Thurstone scores is significantly greater than chance

124

(p < .01) in each case, and in no case is the across-tech-
niques correlation significantly lower than the reliability
coefficient for the Thurstone test . . . It is apparent,
then, that whatever the Thurstone scales measure, the evalu~
ative factor Of the Semantic Differential measures just as
well" (pp. 193~l94).

In comparison with a Guttman scale, the rank order
correlation between it and the 'evaluative' factor Of a
Semantic Differential scale was highly significant (p < .01).
Osgood, Suci, and Tannenbaum concluded that the Guttman
scale and the Semantic Differential scale were, to a consid-
erable degree, measuring the same thing. The semantic
factors have been validated through factor analytic pro-
cedures. "When the intercorrelations among many scales
are factor analyzed and certain basic factors, such as
evaluation, potency, and activity, repeatedly appear, we
assume that these factors correspond to the major dimensions
which people 'naturally' and 'spontaneously' use in making
meaningful judgments" (Osgood, Suci, and Tannenbaum, p. 143).

Osgood, Suci, and Tannenbaum (1957) defined attitude
as primarily the 'evaluative' component of their Semantic
Differential model. They acknowledged that other factors,
'potency' and 'activity' being the most contributive, do
add to attitude, but that evaluation is the prime charac-
teristic. Attitude was defined by Osgood, Suci, and Tannen-

baum (1957, p. 190) as "a learned implicit process which is

125

potentially bipolar, varies in its intensity, and mediates
evaluative behavior." It was viewed as an internal media-
tional (or guiding) activity working between stimulus-
response behavior. This model posits attitude as part of the
semantic structure of an individual. Factor analysis is the
method of choice to ferret out from meaning, the component
that is attitude.

In their research using factor analysis, Osgood,
Suci, and Tannenbaum (1957) have identified the evaluation
component; it has usually been found to be the dominant
factor. "It seems reasonable to identify attitude, as it
is ordinarily conceived in both lay and scientific language,
with the evaluative dimension Of the total semantic space,
as this is isolated in the factorization Of meaningful
judgments" (p. 190).

Osgood, Suci, and Tannenbaum (1957) stated that they
were measuring attitude, which they defined as primarily
the 'evaluative' component Of the Semantic Differential.
It is probable that they indeed were measuring a part Of
attitude, but attitude is a very complex variable composed
of many facets, and it is unlikely that their measure takes
into account all of attitude. Their scale consists mainly
Of cognitive items, and does not attempt to measure any of
the action (conative) component of attitude.

In terms of the Operations of measurement with the

semantic differential, we have defined the meaning
of a concept as its allocation to a point in the

126

multidimensional semantic space. We then define
attitude toward a concept as the projection of this
point onto the evaluative dimension of that space.
Obviously every point in semantic space has an
evaluative component (even though the component may
be of zero magnitude, when the evaluative judgments
are neutral), and, therefore, every concept must
involve an attitudinal component as part of its
total meaning (Osgood, Suci, and Tannenbaum, 1957,
pp. 190-191).

Osgood indexed attitude by using concepts that have
been shown to be high in evaluation. Test-retest reliability
data from various studies using this method has ranged from
between .87~.93; the methods have displayed considerable
face validity. Jenkins, Russell, and Suci (1958) conducted
a normative study of the Semantic Differential in which 360
words were rated on 20 scales by 18 groups Of 30 subjects.
Test-retest reliability for this study was .97. Studies
that have used both Thurstone and Guttman scales lend
evidence to the notion that the evaluative dimension of
Osgood's Semantic Differential is a measure of attitude.
Osgood believed that the evaluation factor of the Semantic
Differential can be used as a generalized attitude scale
(Osgood, Suci, and Tannenbaum, 1957). Since their test
measures primarily the cognitive component, it is question-
able whether it can be considered a general measure of
attitude. Osgood, Suci, and Tannenbaum have no methodologi-
cal Or statistical system to support their theoretical
system; Guttman (1959) felt that to develop a sound theory,

it must have a sound structural basis, as well as a semantic

definitional system.

127

To improve prediction of attitude, other dimensional
scales may be added to the evaluation dimension. "The rela-
tive weights of these factors have been fairly consistent:
evaluation accounting for approximately double the amount
of variance due to either potency or activity, these two in
turn being approximately double the weight of any subsequent
factors" (Osgood, Suci, and Tannenbaum, 1957, p. 325).
Osgood stated that two individuals may have the same atti-
tude toward a certain concept (i.e., Negroes; abortion reform)
as measured by the evaluation factor, but the concept may
have quite different meanings to each individual rating it.
Tannenbaum ran a study where one individual rated the
concept, THE NEGRO, as unfavorable, strong, and active,
whereas another rated this concept equally unfavorable,
but also as weak and passive. NO behavioral data were
available, but Tannenbaum felt that these two individuals
would behave quite differently if put in similar situations.
"While it is true that different attitudes imply different
behaviors toward the Objects specified, at least in some
contexts, it is not true that the same attitude automatically
implies the same behaviors" (p. 190). The greatest crit-
icism Of this conception, was that no systematically measured
behavioral data were used. Clinical Observation alone is
a weak basis for a theory of attitude.

Fishbein (1965) saw Osgood as presenting attitude as

a unidimensional concept. Although Osgood stated that the

128

definition Of attitude included more than merely the eval-
uation factor, Fishbein (p. 108) felt that Osgood's defini-
tion involved only the "evaluative meaning of an Object
or a concept-~its 'favorableness' or 'unfavorableness',
its 'goodness' or 'badness'."

"Fishbein and Raven (1962, cited in Fishbein, 1965,
p. 109) suggested a definition of belief that is analogous
to Osgood's (1957) definition of attitude". Of the six
types Of beliefs (on the following page), in Fishbein and
Raven's definition, the first four involved the cognitive
component or structure while five and six were concerned
with an individual's behavioral intentions or the action
component of belief. Although not stated by these authors,
it appears that the affective component would be involved
in Levels 4 and 5 of the Guttman-Jordan system. Their
definition would then include the cognitive-affective-
conative trichotomy; since they see their definition as
analogous to Osgood's definition of attitude, this would
lend support to the belief that Osgood's definition included
these three elements.

These six types of beliefs have been classified as
follows:

1. Beliefs about the component parts Of the Object;

2. Beliefs about the characteristics, qualities,
or attributes of the object;

3. Beliefs about the object's relations with other
Objects or concepts;

4. Beliefs about whether the Object will lead to
or blOCk the attainment of various goals or
'valued states';

129

5. Beliefs about what should be done with respect

to the Object;

6. Beliefs about what the object should. or should
not, be allowed to do (Fishbein, 1965, pp. 110~

111).

Osgood (1965) ran cross-cultural comparisons on a
semantic differential scale using concepts high on evalua-
tive loadings, indicative of Osgood's definition of atti-
tude. Intercorrelations were performed in seven countries
invOlving one hundred common concepts. On three dominant
factors of evaluation, potency, and activity, Osgood found
that they existed across all cultures that he studied.

He concluded that these factors represent common semantic
dimensions, and are not related to specific cultures.

There exists a common meaning system, panculturally;
individuals use similar symbolic dimensions in organization
of their thoughts and their experiences. Osgood pointed
out that this is one aspect of language that has been found
to be universal. Therefore, this method can be used to
measure 'subjective culture'~~meanings, attitudes, values,
customs~~across different cultures and languages (Triandis,
1972).

Brinton (1969) used Guttman scale analysis in
deriving an attitude scale from Semantic Differential data.
He found this method useful "in testing dimensionality of

the selected adjectives and in ordering individuals on the

dimension" (p. 473).

130

Color Labelling and the Semantic
Differential

 

 

Williams and his associates have been conducting
studies using the Semantic Differential and applying it to
the concepts of race and color. In one Of his earliest
studies, Williams (1964) confronted the question of whether
the words, 'black' and 'white,'could carry a meaning that would
encourage or maintain the learning of race prejudice. He
cited evidence indicating the negative use that the word
or color, 'black,‘ has in our society; in each example,
blackvwnsassociated with something unpleasant and had a
connotation of badness. Some of these symbolic meanings
that Williams (1964, p. 721) cited include: "things look
black, to blacken one's reputation, blackmail, black list,
blackball, black sheep".) Religion and the supernatural
contain much of the same: devils and sin as black, angels
and heaven as white. In all types Of literature from
children's stories to the mass media the same type of
black-white symbolism is found. Researchers have devoted
little time or effort to this black—white word dichotomy.
"The Observations lend support to the generalization that,
in our culture, black symbolizes badness and white sym-
bolizes goodness" (Williams, 1964, p. 722). Jenkins,
Russell, and Suci (1958) found that 'white' and 'light'
were given a positive rating, while 'dark' was evaluated
negatively. Other studies (Fanon, 1967; Podair, 1970;

Triandis 1972) supported Williams' conclusions.

131

In one study, Williams (1964) found a high degree Of
consistency in evaluation Of color names. "Color names
were shown to be quite similar across both geographical and
racial lines providing strong support for the notion that
the connotations of colors are learned via experiences
common to most persons in our general culture" (p. 728).
Both white and black persons evaluated the color name white
more positively than they evaluated the color name black.
Black and white were given quite different ratings on the
three factors Of 'evaluation' (E), 'potency' (P), and
'activity' (A). On both potency and activity, significant
differences were found. The color name black was given a
rating of 'strong' by both black and white subjects; white
was rated as 'weak.' These subjects also rated black as
somewhat 'passive' and white as somewhat 'active.’ These
resultsvmnxgsomewhat different than past results. Usually,
P and Awere slightly positively correlated with E; here,
the results of P and A correlatedixiopposite directions
(Williams, 1964).

Of the ten color names tested, black received the
most negative rating on the evaluation dimension. Williams
(1964) cited a study by Staats and Staats (1958) that
stated that evaluative meanings can be conditioned and
modified by regular association with other words having a
positive or a negative connotation. A laboratory study by
Williams and Edwards (1969) illustrated successful condi-

tioning to the words 'black' and 'white.‘

132

From this, one predicts that the regular associa-
tion of the term Negro with the term black, and
the association Of Caucasian and white, would tend
to condition the connotation of 'badness' to the
former and 'goodness' to the latter. Such an
effect could be a significant background factor in
Caucasian prejudice against the Negro, serving

to facilitate the original learning Of prejudice
in childhood and to support prejudice among adults
(Williams, 1964, p. 730).

The principal conditioning probably occurs in the
above model; however, once prejudice has been established,
some conditioning may occur in the Opposite direction. A
'vicious circle' of mutual reinforcement may then be set
up. This cause-effect relationship is still unclear and
additional studies are needed (Williams, 1964). The impor-
tance of this study lies in the discovery that once preju-
dice is learned, small reinforcements can lead to mainten—
ance of the original prejudice. There is ample Opportunity
in our society to maintain this 'vicious cycle' Of prejudice
once learned in early childhood.

In another study, Williams and Carter (1967) found
further support for the idea that designation of a group
by a color name influences perception of that group by
others. "Triads of concepts linked by the color code (e.g.,
Black-Black person-Negro) were significantly more similar in
connotative meaning than were triads Of concepts not so
linked (e.g., Black-Red person-Oriental)" (p. 19). This

study was repeated in Germany to discover if the findings

were specific to American culture or whether they had

133

greater generality within Western culture. Similar results
were found in the German sample.

Williams and Roberson (1967) found that the develOp-
ment of racial attitude and color meanings occur at about
the same time, with develOpment of racial attitude occurring
slightly sooner. "This result does not support the earlier
hypothesis (Renninger and Williams, 1966) that the black-
white color meanings are learned first, and provide a frame
of reference for the learning of evaluative responses to
racial groups designated as 'black' and 'white'" (p. 687).
Williams and Roberson concluded that the color—meaning
factor acts more as a reinforcing element in the develOp-
ment of prejudice in childhood.

Another cross-cultural study using the Semantic
Differential examined attitudes of five different groupings
of peOple in four societies. Morland and Williams (1969),
using the Semantic Differential developed by Williams in
1964, tested college students representing the following
groupings: American Caucasian and American Negro, Asiatic
Indian, German Caucasian, and Hong Kong Chinese. For the
American samples, white attitudes for this new study proved
similar to white attitudes tested in 1964. The attitudes
of blacks did change. Both words 'Caucasian' and 'American'
were given less favorable ratings in the latest study;
'American' was given a rating much closer to the concept of

'Caucasian' than to the concept 'Negro.' These blacks rated

134

'African' more positively than the comparable sample of
blacks tested in 1964, and rated 'Negro' higher than 'Friend'
in the more recent study. Morland and Williams (1969) con-
cluded "that these indications of changes in attitudes
reflect the feelings growing out of a heightened sense of
racial identity among Negro Americans." The authors noted
that a limitation of their studyvmusthat "the evaluative
factor of the Semantic Differential measures the direction
of the attitude rather than its content" (p. 110).

Williams and Edwards (1969) discussed a laboratory
study involving preschool children and the modification of
color and racial concept attitudes. They attempted to modify
attitudes through laboratory reinforcement conditioning
procedures. The results of the study indicated that the
negative associations of black can be reduced, the positive
association of good with white can be weakened, while not
creating a reversal of the associations with these colors.
The possibilities for use and application of this principle
on a larger scale merits further investigation.

The study can be viewed as providing evidence in
support of the hypothesis concerning a functional
link between the black-white concept attitude and
the racial concept attitude; children whose black-
white concept attitude had been weakened subse-
quently showed somewhat less tendency to evaluate
Negroes negatively and Caucasians positively. The
change in racial attitude attributable to the

experimental treatment, however, was not great
(Williams and Edwards, 1969, p. 748).

135

Williams (1970) tested the relationship of color-
coding practice to the perception of racial groups. Tradi-
tional color-coding practice associated Caucasians with the
color white, Negroes with the term black, Orientals with
yellow, American Indians with red, and Southwest Asians with
the color brown. "Color coding might operate as a background
factor in the development and/or maintenance of attitudes
toward racial groups" (p. 38). Each of the ten color names
studied by Williams (1964) was paired with the word 'person'
(black person, white person, brown person, yellow person,
red person, blue person, green person, purple person, orange
person, and grey person) and evaluated on the three dimen-
sions, evaluation (six concepts), potency (three concepts),
and activity (three concepts). Results for Caucasian
subjects indicated that racial concepts do have connotative
meanings similar to their color names. The most consistent
findings were for the evaluation factor; less consistent
results were found for potency and activity. Williams
stated that this is important since Osgood, Suci, and
Tannenbaum (1957, p. 193) pointed out that "score variation
along the E dimension covaries closely with the score
variation on conventional attitude tests." Caucasian
subjects rated Caucasians most favorably, Asiatic Indians
and Negroes least favorably. This compared with their
evaluative ratings of color names. Williams concluded

that "while the direction of cause and effect cannot be

136

demonstrated here, these data are consistent with the notion
that the evaluative connotations of color names applied to
racial groups are one determinant of the favorability of
attitudes toward the racial groups" (p. 47). Williams
(1969, pp. 384-385) "found that subjects wh0\fifiwmxithe color
name Black more positively had somewhat more favorable
attitudes toward Negro persons, while subjects who viewed
Black more negatively had somewhat more negative attitudes."

A suggestion by Williams (1970) to remove the possible
detrimental effects of this color code-racial concept asso—
ciation would be to eliminate the reference to racial groups
by color names. This, he felt, would eliminate one crucial
channel that has been reinforcing negative connotations.

The variable of voluntariness must be considered. Positive
factors derived by blacks being able to choose their own
'label' (whichever they may desire), may account for
greater gain than any negative loss through association
with negative concepts.

Renninger and Williams (1970) found similar results
in their study of preschool children. At a very early age,
children learn the associations of the word black with
negative concepts and white with positive conceptions.

This can easily generalize to persons or groups of persons
labelled black, brown, white, etc.; hence, the learning of.
the language and affect of prejudice. These authors found

similar results for both black and white children. "As

137

James Baldwin has written (1962, p. 65, cited in Renninger
and Williams, 1970, p. 320),"Negroes . . . are taught to
despise themselves from the minute they open their eyes on
the world. This world is white and they are black." One
task of social scientists can be the association of black
with things positive; blacks have begun to do this with the
"black is beautiful" movement, but whites have done little
to help this cause.

Another study conducted by Williams and McMurtry
(1970), supported the previous work of Renninger and
Williams (1966) and Williams and Roberson (1967). They
found much similarity between seventh grade students'
responses on what they perceived in the affective meanings
of color names and what college students affectively per-
ceived in the color names. They concluded that this affec-
tive perception is operative as early as age thirteen and
probably considerably earlier than this age. "Thus, the
evaluative meanings of white and black appear to develOp
quite early in life and hence are available to influence
the formation of attitudes toward groups of persons desig-
nated, quite inaccurately, as 'white' and 'black'"
(Williams and McMurtry, 1970, p. 713).

A cross-cultural study conducted by Williams,
Morland, and Underwood (1970) in the United States, Europe,
and Asia supported their contention that connotations of

color names were not only evident in the United States, but

138

were also present in different cultures. They stressed
that for American culture, the move to substitute the word
'black' for 'Negro' may not be so wise. The negative
associations with the word black may then be conditioned to
the group of persons referred to as black. They felt that
the word Negro does not have these negative color name
meanings, until it becomes associated with the term black
early in a child's life. However, it would appear that
just as the child learns the negative connotations associated
with the word 'black,‘ he would as easily learn the negative
associations of the word 'Negro' in this society.

Williams, Tucker, and Dunham (1971), in a review of
changes in the connotations of color names among Negroes
and Caucasians, noted that most studies indicated that both
Caucasians and Negro students rated the color name white as
good, weak, and active, and the color name black as rela-
tively bad, strong, and passive. During the mid— and late
1960's, there was a heightened acceptance of the term 'black'
by black Americans; it became a rallying point of identi-
fication. This was one of the reasons Williams, Tucker,
and Dunham replicated the 1963 study in 1969. For black
subjects, the concept 'black' became more positive (evalua-
tion of good) and more active; the conception of strength
remained about the same as it had been in 1963. The term
'white' was rated by black subjects as less positive and

less active; the rating of weakness remained the same. The

139

authors cited the black identity movement as a major cause
of this change. For white subjects, no changes occurred
for either the term 'black' or the term 'white.' Three
other color names were included in both studies, but no
changes were evident for either black or white students.

The events of the years 1963—1969, including the

development of the black identity movement, appear

not to have altered the general Caucasian view of

black as being bad, strong, and passive, and

white as being good, weak, and active . . . Thus,

it would appear that as of 1969, the impact of the

black identity movement had no appreciable effect

upon the meanings of color names to Caucasian

persons (Williams, Tucker, and Dunham, 1971, p. 228).

Another study conducted by Williams, Best, Wood, and

Filler (1971), tested whether the color 'black' had gone
through any changes due to the black power movement, despite
lack of changes in the concept, 'black person,‘ for Cauca-
sians. This study found no significant changes for any of
the color names or ethnic concepts for Caucasian college
students during the 1963-1969 time period. There was some
significant changes in the ratings of the color—person
concepts-~the affective meanings of color-person concepts
became more closely associated with their ethnic concept.

' was more

For example, the color-concept, 'black person,
nearly associated with the 'ethnic' concept, 'Negro,' in
this 1970 study than it was for the 1963 study. Williams,
Best, Wood, and Filler concluded that the color person
meanings were coming closer to the ethnic concept meanings.

Meanings attributed to the ethnic concepts themselves

(i.e., Caucasian, Negro), however, had not been

140

changing. The authors' noted that the black power movement
has probably resulted in an increased acceptance by whites<3f
usage of the terms black for black people and white for
white people. This, they pointed out, was purely a change
in rhetoric, without any meaning changes in the ethnic
concepts, Caucasian and Negro, during the time period from
1963 to 1970.

Lessing and Zagorin (1972, p. 62) used a form of the
Semantic Differential "as an exploratory investigation of
the extent to which the cognitive and emotional restruc-
turing advocated by black power advocates was actually
being accomplished." Two hundred seventy—two white and one
hundred seventy-six black college students in five separate
colleges and universities were tested. Lessing and Zagorin
used a measure of the degree to which students had a high
or low degree of "black power orientation" and a Semantic
Differential measure of attitude toward the following
concepts: 'friend,"enemy,"ideal person,‘ 'colored
person,‘ 'black person,‘ 'white person,’ and 'Negro person.’
Among their eleven hypotheses tested was the assertion that
students high on the measure of "black power orientation"
would score high on the 'evaluation' dimension for the
concept 'black person' on the Semantic Differential and
receive a lower score on the E of the Semantic Differential

for 'black person' than for 'Negro person' or 'colored person.

141

The measure of "black power ideology" used in this

study was a twenty-three item Black Power Ideology Scale,

 

developed by Lessing. One of Osgood, Suci, and Tannenbaum's
(1957) Semantic Differential techniques was used as a measure
of attitude toward various racial groups. Their particular
choice of concepts was based on studies of Williams (1966)
and Williams and Carter (1967). The concepts rated were
as follows: 'friend,‘ 'enemy,’ 'ideal person,‘ 'black
person,‘ 'white person,’ 'colored person,‘ and 'Negro person.’
Each concept was rated on twenty scales, ten representing the
'evaluation' dimension, five representing 'potency,‘ and
five representing 'activity' (p. 62).
One major purpose of this study was to answer the

following question:

If one compares high and low scorers on a measure

of black power ideology focussed mainly upon its

political and economical aspects, is there inde-

pendent evidence that the high pro—black-power

scorers express less depreciation of blacks and

less idealization of whites than the low black power

scorers? (Lessing and Zagorin, 1972, p. 69).
The results of the study confirmed that for both black and
white students, those high in black power orientation gave
'white person' a lower 'evaluation' rating.

Williams (1966) wrote that the connotative associa-

tions with the word 'black' were negative and will probably
remain so; Lessing and Zagorin (1972) felt there have been

positive changes made by youthful black power advocates;

their results indicated that one major effect of the black

142

power movement "was the removal of the odious connotations
of black without giving it clear and unequivocal priority
over other non-white ethnic designations" (p. 69). In
Williams' 1966 study he found that both white and black
students gave thecxxuxxnz'black person' a much lower
'evaluation' than 'Negro person.‘ In Lessing and Zagorin's
(1972) study, those low in black power ideology likewise
evaluated black person lower, but not significantly so,
than Negro person. Those high in black power ideology did
not show a significant reversal of these results. Williams,
Best, Wood, and Filler (1971) found that in 1970, whites
perceived the concepts 'black person' and 'Negro' as more
closely associated than they had in 1963.

Although there were no clear or significant results
for the 'evaluation' factor, there were for the 'potency'
and 'activity' factors. "Blacks were seen as the strongest,
most potent, and most active. Negroes were viewedeusslightly
less strong and active, whites as still less strong and
active, with colored persons being viewed as the weakest
and most passive" (Lessing and Zagorin, 1972, p. 70).

Though the word 'Negro' was once held to have the
same meaning of 'slave' by a pre-Civil War court
(Isaacs, 1963, p. 65), the contention of Baird
(cited in Bennett, 1967, p. 52) that 'Negro' evokes
a slave image in the minds of twentieth century
Americans is questionable on the basis of the
present findings: Negro person was perceived as

no weaker than white person. Clearly, however, the
designation 'black person' was the most evocative
of the powerful image which black power advocates

desire for Afro-Americans (Lessing and Zagorin, 1972,
p. 70).

143

Lessing and Zagorin (1972, p. 72) concluded that
their "research suggests that acceptance or rejection of
black power ideology has verbal attitudinal correlates."

Osgood (1973) has provided an extensive evaluation
of color preferences and connotations across cultural bound-
aries. He had found that BLACKvwnsthe most 'disapproved'
(taboo) color, being so in 33 of 57 countries evaluated;
WHITEvufisthe least disapproved, disapproved in only six of
57 countries. With respect to the potency (P) factor, in an
Asian community, WHITE was considered light (individual's
responses of 8 out of a possible 10), while BLACK was rated
heayy (1 out of a possible response of 10 on a potency
scale). Another study evaluating 17 countries, found bright
colors consistently selected and preferred over dark colors.
In a German sample, it was found that "brightness correlated
positively with a 'happiness' cluster (presumably E+),
negatively with 'forcefulness' (P+) and negatively with
'calmness' (P+ and A-?)" (p. 47). Other studies reported
preferences of WHITE first, BLACK second, and GREY last;
studies in Western European and East Asian communities
reported WHITE rated as 3992 and both GREY and BLACK rated
as bad.

For American English speakers, both BLACK and WHITE
have uniqueness of affect attribution and we suspect
that despite the parenthetical addition of (COLOR)
they carried racial undertones. In the low P but
high A of BLACK, the high CI (Conflict Index, a

measure similar to the E, P, and A measures) of
WHITE, and the fact that WHITE has less A than

144

BLACK (all in contrast to universal trends), one
can perhaps see the impacts of the Civil Rights
Movement, of the Black Panthers and of the rally
cry of 'Black Power' upon our dominantly white
Midwestern small-city teenagers; Williams et. a1.
(1970) confirms the unusually high A of BLACK for
their Caucasian AE White subjects (this sample

was American college students), but not the low P
that we find. The high P and FAM (familiarity) of
COLOR seems consistent with heavy eXposure to ad-
vertising in the mass media and with near—saturation
of color-T.V. (Osgood, in process, p. 70).

In a discussion of universal trends in color pref—
erence, connotation, and association, Osgood (1973) stated
that the only universal for WHITE was that it was low on
the Potency dimension. BLACK and GREYumnxaboth low in
Evaluation and Activity, but differed substantially on
Potency. On Potency, BLACK was nearly P+ and GREY was
clearly P-; BLACK was high in CI and GREY was low in
Familiarity.

Osgood (1973, p. 96) discussed color connotations;

he defined 'connotations' as "operationally-adjectival

characterizations (e.g., adventurous, solemn, cowardly)

 

and emotion-related nominal characterizations which can be

readily transformed into adjective (e.g., valor, female,

 

passion, anxiety, disgust)." He divided color associations

 

 

into four categories: (A) Concrete Identifications, (B)
Concrete Associations, (C) Abstract Association, and (D)
Abstract Symbolisms. The research on the four categories
was done cross-culturally in 37 different regions. This

recent cross-cultural study clearly illustrated that the

145

'color' white was universally associated with positive
concepts and black almost always connoted the negative.
Grey, which is often associated with black, also is viewed
negatively. The Concrete Associations category is the most
realistic of all the categories; the color is the associa-
tion (i.e., white doctor's uniform). The next two cate-
gories, Abstract Association and Abstract Symbolisms, are
much less reality-based. It can be seen that most of the
examples under Abstract Association have little to do with
actual color representations (i.e., white representing
eternity, virtue, innocence, life; grey and black connoting
discouragement, fatigue, sin, falsehood, Satan). Some of
the connotations (one-sixth of those listed by Osgood) of
white do represent negative meaning (mourning, sin, and
death); however, all the concepts listed under Abstract
Association for the 'colors' grey and black were negative.
Osgood (1973) studied 37 different cultural regions; in

only three of these regions was the 'color' white given a
negative 'evaluative' rating. Likewise, for grey and black,
the majority of the regions studied (34 out of 37) gave

grey a negative 'evaluation,‘ along with a stronger 'potency'
rating (i.e., word associations such as sin, death, and
murder). The following analysis of these three 'colors'
for the three categories was taken from Table 6:19 (Osgood,

1973):

146

B: Concrete Associations

 

WHITE: baby, child, daughter, doctor (uniform),
flowers . . .
BLACK: dirt, mud, winter . . .

C: Abstract Associations

 

WHITE: eternity, the future, virtue, innocence,
holiness, chastity, modesty, virginity,
purity, integrity, life, mourning, sin,
death, truth, heaven, light, marriage

GREY: mourning, discouragement, pessimism,
negation, fatigue, boredom
BLACK: sin, mourning, humility, death, murder,

falsehood, negation, Satan, devils, hell

D. Abstract Symbolisms

 

WHITE: Brahman (highest) caste, flag of truce

BLACK: Siva(India), Sudras (lowest) caste
(India)
(1) Concrete color identification--names of

things naturally (or normally) having
a given color (e.g., WHITE--chalk,
BLACK--night)

(2) Concrete color associations--names of
things culturally associated with a
given color (e.g., BLACK-—necktie, RED—-
sacrifice)

(3) Abstract color associations-—terms for
non-"point-at-ables" which can only
have color metaphorically (e.g.,GREY--
Monday, BLUE-—eternity, RED-~patriotism)

(4) Abstract color symbolisms--culturally
significant concepts which certain colors
"stand for" or represent traditionally
but not in any obvious way metaphorically
(e.g., colors associated with castes
in India, with certain deities, religions,
etc.)

 

 

The vast majority of associations reported in
the literature were readily codable as abstract
associations-~that is, concepts which have no color
literally and hence whose associations with color
must be based upon a metaphorical relation, e.g.,
virtue, innocence and chastity have no inherent
color and their frequent association with WHITE
suggests a common affective mediation. In this
sense, we can say that such metaphorically-based
associations are also non-arbitrary.

 

147

With three exceptions (3 cultural communities),
WHITE has E+ associations--virtue, urit , truth,
heaven, and the like; GREY has conSistently E- (as
well as P- and A-) associations--pessimism, fatigue,
boredom, for example; BLACK is equally E-, but
obviously higher P--sin, death, murder, for example,
to say nothing of Satan . . .

 

 

Conclusion

 

Throughout history, 'black' has been associated with
negative words, while 'white' has been associated with posi-
tive conceptions. Many studies (Bennett, 1967; Jenkins,
Russell, and Suci, 1958; Lessing and Zagorin, 1972; Osgood,
1973; Podair, 1970; Schwartz and Disch, 1970; Williams,
1964, 1966, 1969) support this belief.

The Semantic Differential technique was developed by
Osgood (1957) and his associates to measure semantic meanings
of words and concepts. Much of their research involves an
attempt to define attitude, which they see as primarily a
combination of three components, 'evaluation,' 'potency,‘
and 'activity.' Of these three, 'evaluation' is seen as
representing the primary component of attitude. Osgood
(1965) has found that the three concepts, 'evaluation,'
'potency,‘ and 'activity,‘ appear cross-culturally; he felt
that this was evidence supporting the concept that a common
meaning system exists across cultures;individuals use
similar symbolic dimensions in the organization of their
thoughts and their experience.

Williams (1964, 1966) and his associates have

applied the Semantic Differential technique to race and

148

color. Their research concluded that 'black' has been sym-
bolically associated with negative concepts, and 'white'
with quite the opposite. Extensive studies have been con-
ducted by Williams and his associates between the years of
1964 and 1971 (Williams, 1964, 1966, 1969, 1970; Williams,
Best, Wood, and Filler, 1971; Williams and Carter, 1967;
Williams and Edwards, 1969; Williams, Morland, and Under—
wood, 1970; Williams and McMurtry, 1970; Williams and
Roberson, 1967; Williams, Tucker, and Dunham, 1971; Morland
and Williams, 1969; Renninger and Williams, 1970) to support
these conclusions; additional studies have been conducted
using and/or citing the work of Williams (Lessing and
Zagorin, 1972; Osgood, 1973; Triandis, 1972). Jenkins,
Russell, and Suci (1958) found that 'white' and 'light'
were given a positive rating, while 'dark' was evaluated
negatively.

Williams and Edwards (1969),in a laboratory experi—
ment, found that the negative associations of black can be
reduced, the positive associations of white can be weakened,
without creating a reversal of the associations with these
colors. Their subjects showed less tendency to evaluate
Negro negatively and Caucasian positively. The authors
found a functional link between the colors, black and
white, and the racial names, Negro and Caucasian.

Williams (1970) has suggested that since he found an

association between color-coding and racial concepts (i.e.,

149

the negative connotations of black become associated with the
racial group 'black'), it would be beneficial to eliminate
the reference to racial groups by color names. The present
racial ABS study, however, found no differences between

the words, 'black' and 'Negro', among a group of college
students. This would indicate that Williams' above sugges-
tion might have been premature.

The results of these studies indicated that although
there once was a distinction between the concepts black
(black person) and Negro (Negro person), this distinction
probably has begun to dissipate. Williams' earlier studies
(1964, 1966) found a distinction; two of his latest published
studies (Williams, Best, Wood, and Filler, 1971; Williams,
Tucker and Dunham, 1971) indicate lack of a difference.
Williams, Best, Wood, and Filler (1971) found that color
person meanings are coming closer to ethnic concept meanings,
i.e., 'black person' was more nearly associated with 'Negro'
in a 1970 study than was found in 1963. They pointed out
that the recent black power movement has probably resulted
in increased acceptance by whites of the term 'black' for
black people. This conclusion was supported by the present
ABS study and also by Lessing and Zagorin's (1972) study.
Although they found some slight differences, they concluded
that 'Negro' no longer has the negative connotations it

might have had sometime in the past.

CHAPTER V

METHODOLOGY

To satisfy our doubts, . . . therefore, it is
necessary that a method should be found by which
our beliefs may be determined by nothing human,
but by some external permanency-darsomething upon
which our thinking has no effect . . . The
method must be such that the ultimate conclusion
of every man shall be the same. Such is the
method of science. Its fundamental hypothesis . . .
is this: There are real things, whose characters
are entirely independent of our opinions about
them . . . (Buchler, cited in Kerlinger, 1964,

p. 7).

The methodology for this thesis is based on Jordan's
extensive research at Michigan State University. Jordan
has expanded and refined Louis Guttman's four-Level facet
theory into a six—Level design, maintaining the simplex
structure.

Jordan (l971b) applied his new five-facet, six-Level
research design to the mentally retarded in seven nations.
In reviewing the literature in the area of attitudes toward
mentally retarded persons, Jordan (1970) found no studies
that had employed a Guttman facet design (p. 3). Thus,
attitude research in this area using a facetized design is
a relatively new and recent innovation.

Out of this extensive research on attitudes toward

the mentally retarded, Jordan developed the Attitude Behavior

 

150

151

Scale toward mentally retarded persons (ABS-MR). Further
research and investigation lead to modification of the
scale, and also to the development of parallel scales for
groups other than the retarded. The need for more concise
and systematic measurement of attitudes is cited by Jordan
(l971a):

Extensive reviews (Bray, 1970; Campbell and Schuman,
1969; Dell Orto, 1970; Erb, 1969; Frechette, 1970;
Harrelson, 1970; Morin, 1969; Poulos, 1970; Robinson
and Shaver, 1969; Robinson, Rusk and Head, 1968;
Vurdelja, 1970; Williams, 1970) of the studies of
attitudes toward such diverse attitude objects, as
mental retardation, deafness, blindness, mental
illness, war, religion, nationality, perceived
racial differences, indicated that none of the pre-
vious studies employed an attitude scale constructed
on the basis of the structural theory proposed by
Guttman in 1959. It is unclear what attitudinal
levels or sub-universes were being measured in

most, if not all, of these studies; although a
semantic analysis indicates that most of the scales
fall at the stereotypic level in Guttman's paradigm.
If the latter statement is correct, it may eXplain
the recurring theme throughout attitude research
that "attitudes do not predict behavior" (p. 8).

 

These attitude-behavior scales have not only been
applied to numerous 'minority' groups, but have been trans—
lated into different languages and used cross-culturally.
Prior to Jordan's research, attitude scales had not been
used extensively across different cultures and none were
found in disability. A review of the literature in 1965
failed to indicate even one study "which attempted to
relate findings cross-culturally or cross-nationally . . .
Only three studies (Laing and Chazan, 1966; Schonnell and

Watts, 1956; Schonnell and Rorke, 1959) were found in the

152

American literature which were conducted in countries other
than the United States" (Jordan, 1970, p. 4).

For rigorous methodology to be useful, it must be
based on theory that is relevant to the real world. Jordan's
theoretical methodology and five-facet, six-Level attitude
scale relates closely to McGuire's (1969) cognitive-
affective-conative (knowing, feeling, acting) analysis of
the human condition. Jordan's first two Levels (Table 17,
Levels 1 and 2), specifically involve the cognitive compo-
nent of behavior. This parallels McGuire's formulation of
‘the cognitive aspect of attitudes (also called by McGuire
'perceptual, informational, or stereotypic component') in
which he states "how the attitude object is perceived, its
conceptual connotation--it is the 'stereotype' the person
has of the attitude object" (p. 155). This is the Level
that Jordan feels most attitude studies have dealt with,
these studies neglecting almost entirely the affective and
conative areas. The third Level in Jordan's attitude
'hierarchy,‘ involves the affective component of attitudes.
According to McGuire, this component of attitude (also
called the 'feeling or the emotional component') deals with
the person's feelings of liking or disliking about the object
of the attitude (p. 155). Levels 3 and 4 on the ABS involve“
a combination of the affective and conative components,

Level 5 deals with the affective component, while Level 6

is concerned strictly with measures of conative behavior.

153

McGuire states that "the conative ('action, behavioral')
component of attitude refers to the person's gross behavioral
tendencies regarding the object" (p. 156).

In developing the Attitude Behavior Scale, Jordan

 

has modified the structural theory of Guttman by including
philosophical and theoretical eXplanations of human behavior.
With these theoretical and methodological tools, Jordan has
developed and expanded the range of attitudinal research.
Because of the ABS instrument, it has been possible to
analyze the facets of attitude and thus provide a more

rigorous, methodological framework for attitude research.

Review of Attitude Scales

 

Bogardus (1925) was one of the early researchers to
develop attitude scales. His 'social distance scales'
attempted to measure the amounts, degree, and feelings
associated with social contact that whites perceived them-
selves as having with blacks. Taylor (1971) states that
this type of scale is rarely used today because many
researchers feel it does not give a true measure of racial
attitudes. Scales of this nature have been used by Maus-
tuscelli.(l950)and Proenza and Strickland (1965).

Thurstone's (1931) attitude instruments used a
method whereby judges evaluated the items to be selected.
His racial scales contained almost exclusively stereotypic

items (Frechette, 1970). This method has been criticized

154

for the process used in selecting the item continuum, the
'objectivity' of judges, and the time and effort required
for constructing and scoring of the scales (Hovland and
Sherif, 1952; Jahoda and Warren, 1966; Frechette, 1970).
Variations of the Thurstone scale were developed by Hinckley
(1932) and Rosander (1937).

Other early scales designed to measure attitudes
were developed by Likert and Murphy (1938). They designed
three scales to measure controversial issues of that time,
an 'Internationalism Scale,’ an 'Imperialism Scale,’ and a
'Negro Scale.’ These three measures of attitude were found
to be highly correlated with each other and appeared to
discriminate between 'racists,' 'moderates,‘ and 'liberals'
on the 'race question' of that time period (Taylor, 1971).

The method of intuition was used by Likert (1932)
to select items for use in an instrument. An item was
selected if it was judged to be 'favorable' or 'unfavorable;'
items not in one of these two categories were thrown out.
Likert developed a scale to assess attitudes toward blacks.
Although most items were of a stereotypic nature, some
items could be considered at the hypothetical action level
(Frechette, 1970). Several social scientists (Steckler,
1957; Greenberg,196l; and Campbell and Schuman, 1968) used
modifications of Likert's technique. "Likert's instrument
has been criticized for yielding only ordinal scale data

and items selected by intuition" (Taylor, 1971, p. 15).

155

Another disadvantage of this technique is that often the
total score of an individual has little clear meaning, since
many patterns of response to the various items may produce
the same score (Warren and Jahoda, 1966).

Guttman developed unidimensional scalogram analysis;h
he modified his techniques of attitude measurement with
his introduction of two new approaches, facet design and non—
metric analysis. Guttman-type scales have been developed
to study racial attitudes by Harding and Hogrefe (1952),
Kogan and Downey (1956), Triandis, Levin and Loh (1966),
and Campbell (1968).

Johoda and Warren (1966) criticized Guttman's tech-
nique as being unidimensional, and therefore not capable
of measuring complex attitudes. They also feel that a
scale may be unidimensional for one group of individuals,
but not necessarily for another. Guttman's latest con-
tributions to scale construction and attitude measurement
(i.e., facet design and nonmetric analysis) avoids many of
the prior criticisms since they are multidimensional in
nature and also include an a priori method of item construc-
tion.

Taylor (1971) reviewed new develOpments and changes
in racial attitude research during the 1960's. Greenberg
(1961) devised a scale to measure attitudes and problems
occurring during integration. In an exploration of racial

attitudes, Weiss (1961) developed an instrument to reveal

156

the stereotypes whites have of blacks and an opinion ques-
tionnaire to test dimensionality of racial attitudes.
"Weiss concluded that the distinctions between descriptive
and attitudinal traits implied dimensionality of racial
attitudes" (Taylor, 1971, p. 20).

Guttman's Theoretical Structure:
A Four-Level Theory

 

 

Guttman sees a need to build systematic theories in
the social sciences to provide a basic framework for research-
ers. In the area of attitude research, he has attempted to‘_
provide an abstract framework by defining a measurable set
of sub-universes underlying a structural analysis of inter-
group behavior. Although Guttman based his theoretical forélfl
mulations on interracial behavior, he states that his system
can easily be extended to other groups (Guttman, 1959,

p. 319).

Guttman's (1959) theory involves first a discussion
of semantic structure and second, a statistical analysis of
that structure. He feels that both analyses are necessary “7
to relate more abstract conceptions to actual observation
and experience.

From research completed by Bastide and van den
Berghe (1957), Guttman abstracted definitions for the four
sub-universes of attitudes. Because Bastide and van den
Berghe's work dealt with interracial attitudes, the defini—

tions involve whites and Negroes. The definitions were

reordered by Guttman, and read as follows:

157

l. Stereotype: Belief of (a white subject)
that his own group (excels-
does not excel) in compari-
son with Negroes on (desir-
able traits).

2. Norm: Belief of (a white subject)
that his own group (ought—
ought not) interact with
Negroes in (social ways).

3. Hypothetical Interaction: Belief of (a white subject)
that he himself (will-will
not) interact with Negroes
in (social ways).

4. Personal Interaction: Overt action of (a white
subject) himself (to-not
to) interact with Negroes
in (social ways) (Guttman,
1959, p. 32).

In delimiting the differences in each of the defini-
tions, Guttman notes differences in three facets. "Each

definition concerns a type of behavior of a subject vis—a-vis

 

a type of intergroup behavior of a type of referent" (Gutt-
man, 1959, p. 320). Each of these three facets in turn
has two elements and, therefore, each is dichotomous as

depicted in Table 12.

TABLE 12.--Facets on which Subuniverses Differ

 

(A) (B) (C)

 

Subject's Referent Referent's Inter-

Behavior group Behavior

al belief bl subject's cl comparative
group

a2 overt action b2 subject c2 interactive

himself

 

158

One element from each and every facet must be rep-
resented in any given statement, and these statements can
be grouped into profiles of the attitude universe by multi—
plication of the facets AxBxC, yielding a 2x2x2 combination
of elements or eight semantic profiles in all, i.e.,

(1) alblcl' (2) alblc2 . . . (8) a2b2c2. It can be seen
that profiles 1 and 2 have 2 elements in common (albl) and
one different (cl and CZ), whereas, profiles 1 and 8 have
no elements in common.

The capital letters A, B, and C depict the three
facets, while the subscripts denote the respective elements.
Thus alblc2 reads: Belief (al) of a subject that his own
group (bl) interacts (c2) with a Specified attitude object.
Similarly, a2b2c2reads: Self or observed reports of a sub-

ject's overt action (a2) of himself (b2) interacting (c

2)

with a specified attitude object.
The four subuniverses (Levels) that Guttman derived
from Bastide and van den Berghe (1957) have been facetized

as shown in Table 13.

TABLE l3.--Guttman Facet Profiles of Attitude Subuniverse

 

 

Subuniverse Profile
1 Stereotype alblcl
2 Norm alblCZ
3 Hypothetical Interaction aleCZ
4 Personal Interaction a b c

2 2 2

 

159

There is a rank ordering of facets present in the
design; Guttman refers to it as a progression from a weak
to a strong form of the subject's behavior vis-a-vis the
attitude object. The more subscript '2' elements a profile
contains the greater the strength of the attitude. Also,
there is a progression down the subscale Levels, "stereotype"
(Level 1), being the weakest, proceeding through to
"personal interaction" (Level 4), the strongest i.e.,
l<2<3<4 moving down the Levels from weakest to strongest.

Facet analysis of the semantic structure provides a
social psychological basis for predicting the structure of
the empirical intercorrelation matrix of the above four
Levels.

One cannot presume to predict the exact size of each
correlation coefficient from knowledge only of the
semantics of universe ABC, but we do propose to pre-
dict a pattern or structure for the relative sizes
of the statistical coefficients from purely semantic
considerations (Guttman, 1959, p. 324).

This prediction was stated by Guttman (1959) as
the Contiguity Hypothesis: "Subuniverses closer to each
other in the semantic scale of their definitions will also
be closer statistically" (p. 324). Thus, the Contiguity
Hypothesis predicts that the Levels that are adjacent to one
another will correlate to a stronger degree than will
Levels that are more distant from each other. In other

words, "Normative Behavior" (Level 2) will correlate more

highly with an adjacent Level, "Hypothetical Interaction"

160

(Level 3) than it will with "Personal Interaction" (Level 4),
a more distant Level.

Table 14 illustrates the data obtained in Bastide
and van den Berghe's (1957) study. The data have been
rearranged by Guttman to correspond to his four-Level
ordering, (Stereotype, Norm, Hypothetical Interaction,
Personal Interaction). The structure of the table follows

the simplex pattern, except for one reversal (r .49 does

4.3:

not exceed r .51). This slight error, Guttman states,

4,2:
could simply be due to sampling error and/or sampling bias;
he does not see this error as contradicting the Contiguity
Hypothesis. Further research validates this conclusion by

Guttman ". . . research by Hamersma (1969) and Harrelson

(1970) established a criterion of six reversals as being

 

the maximum that could exist in a six by six matrix and
still permit the simplex to be regarded as 'approximated'"
(Dell Orto, 1970, p. 56). Guttman (1954b) also states that

in actual practice, a perfect simplex is not to be expected.

TABLE l4.--Empirical Intercorrelation of Scores on the Four

 

 

Subuniverses
l 2 3 4
Hypothetical Personal
Subuniverse Stereotype Norm Interaction Interaction
1 Stereotype -- .60 .37 .25
2 Norm .60 -- .68 .51'
3 Hypothetical .37 .68 -- .49
Interaction
4 Personal .25 .41 .49 --

Interaction

 

161

Hamersma (1969) explains that the 'simplex' is
analyzed to ascertain the relationship among the Levels.'
As one proceeds down the Levels, the facets change; however,
to enable the simplex ordering to be possible only one
facet is changed for each successive Level change. A
simplex is defined by Guttman (1954-55) as "sets of scores
that have an implicit order from 'least complex' to 'most

complex (p. 400). In a simplex matrix the ordering of
the Levels (correlations, etc.) is predicted, but the
intensity or magnitude of the correlation is not. Table 15
presents an example of a hypothetical correlation matrix
with a simplex structure, designed by Hamersma.

TABLE 15.--Hypothetical Matrix of Level-By—Level Correlations
Illustrating the Simplex Structure

 

 

Level 1 2 3 4
l _..
2 .60 --
3 .50 .60 --
4 040 .50 050 --

 

Note: One does not attempt to predict the magnitudes
of each correlation coefficient. The simplex requirements
do not necessitate either identical mathematical differences
among various correlations or identical correlations between
adjacent levels, so that the bottom row of the matrix reading
from left to right could contain such figures as .10, .32,
and .49.

Guttman (1959) suggested that to increase the pre-
dictability of his theoretical model, it would be beneficial

to (a) enrich the facet design and (b) place these behaviors

(Levels) in a broader context. Jordan's five-facet, six-Level

162

design encompasses the merits of both these suggestions. He
enriched the facet design by adding two additional facets
and, hence, two additional Levels of behavior. By including
facets that demonstrated more of the affective and conative
dimensions of behavior, he placed the theory in a broader,
feeling-action oriented context, encompassing Guttman's

second suggestion.

Jordan's Six-Level Adaptation

 

Jordan's five-facet, six-Level theory encompasses
Guttman's three-facet, four-Level design, expanding the
theory in the affective and conative domain. Specifically,
Jordan maintains Guttman's four original Levels, but adds
two new Levels toward the lower end of Guttman's scale.

For a visual explanation of Jordan's six-Level theory see
Tables 16 and 17. To compare Jordan's facet system with that
of Guttman's, compare Tables 16 and 17 with Tables 12 and
13. Guttman included four attitude dimension categories:
Stereotype, Norm, Hypothetical Interaction, and Personal
Interaction (Table 13). According to McGuire (1969),
these facets are primarily concerned with cognitive and
affective behavior. Only the last Level, Hypothetical
Interaction, includes any conative material. It is at
this point that Jordan visualized the need to expand
Guttman facet theory. Jordan places more emphasis on the

affective and conative elements of attitude-behavior. His

163

theory, while including Guttman's four Levels (cognitive
and affective elements), extends Guttman into the realm of
conative behavior. His two additional Levels, Personal
Feeling (Level 5) and Actual Personal Action (Level 6)
extend the theory to 'real,‘ observable overt behavior.
These Levels are evaluating the subject's actual feelings
and actions, instead of his perceived thoughts, beliefs,
and opinions (as measured in the first four Levels). They
appear to be the crucial Levels at which attitudinal change
occurs. Tables 17 and 18 contain a more explicit presen-
tation of Jordan's six Levels and a comparison of Guttman
and Jordan facet designations.

TABLE l6.--Jordan Facets Used to Determine Jointa Struction
of an Attitude Universe

 

 

(A) (B) (C) (D) (E)
Referent Referent Actor Actor's Domain of
Behavior Intergroup Actor's
Behavior Behavior
a others b belief c others d comparison e hypo-
l l l .
thetical
a2 self b experience c self d interaction e Opera—
(I) (overt behavior) (mine/my) tional

 

aJoint struction is operationally defined as the ordered sets
of the five facets from low to high (subscript 1's are low) across all
five facets simultaneously.

164

TABLE l7.--Joint Level, Profile Composition, and Labels for
Six Types of Attitute Struction

 

 

Subscale . _ a

Type-Level Struction Profile Descriptive Joint Term
1 al bl cl dl el Societal Stereotype
2 al b1 01 d2 el Societal Norm
3 a2 bl C1 C12 el Personal Moral Evaluation
4 a2 bl c2 d2 el Personal Hypothetical Action
5 a2 b2 c2 d2 el Personal Feeling
6 a2 b2 c2 d2 e2 Personal Action

 

aBased on facets of Table 16.

Tables 18 and 19 propose a structioned definitional
or semantic system for the relationships between the six
scale Levels. According to Jordan (l971a), the Cartesian
product of the five two-element/facets of TableiHSyields 32
possible profiles (Table 21). As shown in Table 19, six
of these profiles were chosen as psychologically relevant,
potentially capable of instrumentation, and possessing a
specific relationship among themselves--a simplex relation-
ship. Maierle (1969) presents an extensive discussion of
the 32 profiles, the specific rules by which the 12 profiles
in Figure 1 are retained, the rationale for choosing six
of these 12 profiles for the six Levels, and the seven
"semantic paths" possible between these 12 profiles; i.e.,
the six Levels presented in Table 19, agree with Maierle's

semantic path C although they were extant prior to that.

165

am no N "m umomm “M_uo M. "Q umomm “m Ho m "U uwomm

mum NH mHQme mo mcoHumcHnfioo

«M.Ho m "m uwomm

um.uo m "a uwomm

"mm mHQMB mo mpcwEmumum HmHusmumMMHp on» cH pmmmmumxm >HHmoHucmew

may msoHHOM mm pwmmmumxm mum 0H oHnme mo mumomm on» me

 

 

 

 

 

 

 

 

 

 

 

 

MHmmEH: m
In o>HuomumucH mp pomhndm o coHuom pum>o ma nu
msoum
II m>HumummEoo Hp m.uomflndm Ho mmHHmn HQ 1:
H0H>mcmm u0a>m mm
In QsoumuwucH ucwummmm . n In
m.uomm35m Illllll
m.ucmummmm ”cmfiuusu
AHOH>man
A>E\mcHEV unm>ov HHV
HmcoHpmummo mm coHuomumucH Np wme No wocmHummxm Nb MHmm mm
H H H H H
HMUHumnuom>£ w somHHmmEoo p muwnuo o mmHHmn n mumzuo m
u0a>m w m no 0 uoH>mgmm u0a>m m
. n m . u < maoumumucH Moved . n m ucmummwm
mo :HmEoo ucmummmm
. m.uouo< ”cannon
m D U m d
coHumcmHmmo

coHumummpa canyon CH mmumomm

 

.mcoHumcmHmmo uwomm canyon paw cmauuso mo COmHummEouul.mH mqmde

166

.mmHHmonm umHHomQSm ppm mumomm Mom hH mHQma mmmo

.mucoEmumum HmcoHQHchmp mom mm mHQdB mwmQ

.oH mHQMB mo mumomm so pmmmmm

 

 

coHuom HMGOmHmm mm mp mo NQ mm m a E m a w

mcHHmmm HMGOmumm Hm mp No NQ mm Q H E m H m

soHuom HMUHquuom>Q HMCOmHmm Hm mp mo HQ mm Q H E Q H v
COHumsHm>m Hmuos HMQOmuom Hm mp Ho HQ mm Q H 0 Q H m
Enos HmuwHoom Hm mp Ho HQ Hm Q H 0 Q o m

mmmuomumum HmumHoom Hm H0 H0 HQ Hm Q o 0 Q o H

Emma mm wHQme :H Emum>m pH mHQme CH Emuwwm mwmmw

m>HumHuome Hm>mQ mpsuHuud QHmcoHpHchwo >Q mHHmoum UHMGOHQmpoz >Q mHHwoum mHmoszm

 

m

.coHuosuum mpsuHuum

mo mmmwa me new mHQOq cam .coHpHmomEoo mHHmoum .Hm>wH ucH0hll.mH mqmde

167

Maierle (1969) states that only 12 of the 32 possible
combinations appear to be semantically and logically consistent.

Construction of the six level Attitude-Behavior Scale:

 

Mental Retardation (ABS-MR) used in Jordan's research was

 

guided by the six combinations of facet elements shown in
Table 17. The six profiles were chosen from the twelve
potential combinations because they appeared potentially
capable of instrumentation, the relevance implied in the
six Levels or 'types,‘ and because they form a simplex
order. The six ABS-MR sub—scales range on an abstract-
impersonal to concrete-behavioral continuum-~from a stereo-
typic Level to reports of actual behavioral interaction
with the retarded; correspondingly, each profile moving
from Level 1 to Level 6 is characterized by one additional
strong facet element in a range from no strong elements to
all strong elements, as is illustrated in the definitions
and examples in Table 17.

It is the intention of the following definitions
and examples to make clear Jordan's six-Level progression
from the cognitive through the affective domain, and finally
arriving at the conative-action Level of attitude-behaviors.
For a graphic representation, including profiles, see

Table 22.

168

TABLE 20.--Permutations of Five Two-element Facetsa of

 

 

 

Table 16.
Facetsb
Permutations

A B C D E

1 1 1 l 1 l
2 l l l 2 1
3 2 l 1 1 1
4 2 l l 2 l
5 1 l 2 l l
6 1 1 2 2 1
7 2 1 2 l l
8 2 1 2 2 1
9 1 2 l 1 l
10 l 2 1 2 1
11 2 2 l l 1
12 2 2 l 2 1
l3 1 2 2 1 1
14 l 2 2 2 1
15 2 2 2 l 1
l6 2 2 2 2 1
l7 1 l 1 1 2
18 l l 1 2 2
19 2 l l l 2
20 2 1 1 2 2
21 l l 2 l 2
22 l l 2 2 2
23 2 l 2 l 2
24 2 l 2 2 2
25 1 2 1 1 2
26 1 2 l 2 2
27 2 2 l 1 2
28 2 2 1 2 2
29 1 2 2 1 2
30 1 2 2 2 2
31 2 2 2 l 2
32 2 2 2 2 2

 

aSubscript "1" indicates weak element; "2" indicates
strong element.

bSee Table 16 for facets.

169

TABLE 21.-—Combinations of Five Two-element Facetsa and Basis of

 

 

 

 

Elimination.
Combinations Facets and Subscripts . c
Ba51s of
b In In Elimination
NO' Table 16 Table 17 A B C D E
l 1 Level 1 o b o c h
2 2 Level 2 o b o i h
3 3 -- i b o c h
4 4 Level 3 i b o i h
5 5 -- a b m c h
6 6 -- o b m i h
7 7 -- i b m c h
8 8 Level 4 i b m i h
9 - -- o e o c h 2
10 9 -— o e o i h
11 - -- i e o c h 1 2
12 - -- i e o i h 1
l3 - -- o e m c h l 2
14 - -- o e m i h l
15 - -- i e m c h 2
16 10 Level 5 i e m i h
17 - -- o b o c p 3 4
18 - -- o b o i p 4
l9 - -- i b o c p 3 4
20 - -- i b o i p 4
21 - -- o b m c p 3 4
22 - —- o b m i p 4
23 - -- i b m c p 3 4
24 - -- i b m i p 4
25 - -- o e o c p 2 3
26 ll -- o e o i p
27 - -- i e o c p 1 2 3
28 - -- i e o i p l
29 - -- o e m c p l 2 3
3O - -- o e m i p 1
31 - -- i e m c p 2 3
32 12 Level 6 i e m i p
aSee Table 16 for facets.
bNumbering arbitrary, for identification only
cLogical semantic analysis as follows:
Basis 1: an "e" in facet B must be preceded and followed by equivalent
elements, both "0"; or "i" in facet A or "m" in facet C.
Basis 2: a "c" in facet D cannot be preceded by an "e" in facet B.
Basis 3: a "c" in facet D cannot be followed by a "p" in facet E.
Basis 4: a "p" in facet E cannot be preceded by a "b" in facet B.

d .
See text for rationale.

.mqusos Hm>mH muoHum> .
mo mmHchoHuMHmu wumoHan mmmwQucmumm :H moan: mumcumuHad Ho>mq cH mucwEon mcouum mo quaac u .OZQ

.hH and 9H meQMB .mpo
.wmc oQu :H pom: mGOHuncHano«¢

.mucmamumum
ucmumHmcou unQ uamvznvou mo uumm mum mmmwQucmHmm 3H mpuosu

1770

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1‘ Q
mHHmcoHumummo AuoH>waQ m ~w~v~o~QNw w o
coHuum Hmcofimm . .3. uum>ov mcoHuumumucM a mocmHummxw .Hu NH ImJHIHeIwIH.) 1
Ill- 1 (11!-
. I AHHmuHquuomzm e HmmpwomQNm m
wcHHmmw HmCOmHmm «s AmwcHHomwv mcoHuumumucH hm mucwHmexo H oH Q H E m H
I- $333533-
coHuum msouw Hmsuo< wcomHHmaeom mHmQum oucmHuuaxM mHmQum HH a H o m o
:oHuum | i l I m mepmoHQNm ¢
..nl.wmoHquuoa>Q Hmsomumm *JI:;WWWWWWWWMWQmmm.mmmwmums mcH HE m>mHHmQ H w IHflwfldflMflJW.
AmwcHHowu aaoum "
Hmsuumv zuHucmpH macho AQQmoHquuoahm _
HmcoHumuuadxo mcoHuomHyusm mHoQum mu:uHummxm mueQum a Q H o m o
msouwv msmH poeHmHUCHm AHHcoHquuoqam
AmSQmum chomuma pocmHmmm moHooQu0de mccHWUmuousm a? s>mWHmm munMum c Q H E Q o
mHHQCOmumav ummucoo-HH¢m AHHmoHquuoaHQ mCCmHummeom we o>mHHoQ H N H Q o E Q H _
AmmsHm> pm>Hmuummv I HHHMUHuwQuomMm _ HoNpHoHQNm m m
coHQmsHm>m Hmuoe QHCOmHmm %% m:1_buwpmucH nu;:uo y>oHHom H , c Q H 0 Q H .
x l I! - .ll-) 4) H
_ ,
msumum _
HmQOmumm ptcmHmmmnasouc xHHmoHquHOQAm madmmumasom hfl z>eHHcm mLcQum _ m Q o E Q o _
Enos HmuwHoom ** >HQmoHquuoaxm V QmmpHoHQHm
msumum mcoHoomuuecw .QOQem o>oHHom mHmQum H M m Q A 0 Q o H N
mmmuw pmcmHmmmumHHm:0mumm AHHmoHueQuoahm mcchpdeom .mpoQum o>oHHom M M m Q o 0 Q H.#
Amsumum _ fl 1
asouw pmcmewm msouwv xHHmoHuwQuoahm W HmeHuHQHm m
mazuompmum HmumHuom ** mCOmHHdeom .meQum w>mHHmm mquum W o w H Q o 0 Q o J H
t -
oEmz m>HumHuome ucmsmumum HmcoHuHcmeo . .OZIQ HNmH as w H cum . m>
. . cH .oz _ umomm .

 

 

 

HmcoHanmeQ can

.mcoHumcHQEoo m>H039 now mandamumum

.mmHHHoum vouch .mHm>mH mumcoHumuHHaQum> mpsuHuud mo Soummm Ho>qume umoumlm>Hmnn.mm mqm<e

171

THE COGNITIVE DIMENSION OF ATTITUDE-BEHAVIOR

Level 1
Societal Stereotype (a Other whites
believe others hypotheticdlly cdmpare. to Negroes.

Directions: Other Whites believe the following things
about Negroes as compared to Whites:

Eg.: Negroes can be trusted with money

(1) less than Whites
(2) about the same as Whites
(3) more than Whites

Level 2
Societal Interactive Norm (a1 b1 Cl d2 e1 ). Other whites
believe others hypothetically (should) interact with
Negroes.

Directions: Most Whites generally believe the following
about interacting with Negroes:

Eg.: Whites believe they can trust Negroes with
money

(1) disagree

(2) uncertain
(3) agree

THE AFFECTIVE DIMENSION OF ATTITUDE-BEHAVIOR

Level 3
Personal Moral Evaluative (a I believe
other whites hypothetically Ishould)d1nteract with
Negroes.

Directions: In respect to Negroes, what do you believe
others think is right or wrong:

 

Eg.: To trust Negroes with money is

(1) usually wrong
(2) undecided
(3) usually right

172

1
THE AFFECTIVE AND CONATIVE DIMENSIONS OF ATTITUDE-BEHAVIOR

Level 4
Personal Hypothetical Behavior (a2b 1 c2 d2 I
believe I hypothetically (would) interact with). Negroes.
Directions: In respect to a Negro person, would you
yourself:
Eg.: I would trust Negroes with money
(1) no
(2) undecided
(3) yes
Level 5

Personal Feelings (a

. I experience (affect)
when I hypothetically ifitegacf with (think about) Negroes.

Directions: How do you actually feel toward Negroes:

Eg.: When Whites trust Negroes with money I feel

(1) bad
(2) indifferent
(3) good

THE CONATIVE DIMENSION OF ATTITUDE-BEHAVIOR

Level 6
Personal Action (a2

I have eXper1ence
myself in operationallyC {actua11y) interacting with
Negroes.

Directions: Experiences or contacts with Negroes:

Eg.: I have trusted Negroes with money
(1) no
(2) uncertain
(3) yes

A mapping sentence is used to provide a concise,

operational definition for the study. Figure 1 represents

 

1The two Levels encompassed under this category

(Levels 4 and 5) include both affective and conative elements
of attitude-behavior.

173

the a priori use of facet theory in the development of an

Attitudes Toward Specified Persons Scale. The mapping sen-

 

tence for the first scale developed by Jordan, Attitude

Behavior Scale--Mental Retardation (ABS-MR) is illustrated

 

in Figures 2 and 3. The mapping sentence that the present

study is based on, Attitude Behavior Scale--Black/White

 

(ABS—B/W) is shown in Figure 4.

Attitude-Behavior Scales (ABS)

The instrument employed in this thesis was an out-
growth of Jordan's attitude-behavior research. Construction
of the scales was based on facet theory and construction
of the items followed a systematic a priori method instead
of by the Likert method of intuition (i.e., even though
the Likert procedure uses item analysis after the initial
selection) or by the Thurstone use of judges. Guttman's
(1959) facet theory specifies that the attitude universe
represented by the item content can be substructured into
behavioral profiles which are systematically related according
to the number of identical conceptual or semantic elements
they hold in common. The substructuring of an attitude-
behavior universe into facets and elements facilitates a
sampling of items within each of the derived profiles and
also enables the prediction of relationships between various
profiles of the universe. This should also provide a set of

Clearly defined profile areas for cross-cultural comparisons.

174

In constructing the first scale, ABS-MR, Jordan
(1970) "postulated that attitudes involve not only objective—
specificity but situation-specificity and object-subject

relationships" (p. 48). In the case of the MR scale, the

 

object was the mentally retarded, situations included such
areas as experiences, education, personal characteristics,
and relationships between the object and the actor (self or

others).

Guttman (1959) suggests a common semantic meaning
for the five facets of Table 16; a progression from a weak
to a strong form of behavior of the subject Vis-a—vis the
attitude object--in this case, the mentally retarded. Exam-

ination of Table 16 indicates the rationale of this ordering

system.

Facet A--the referent 'other' is weaker than 'self'
(I) in being less personal.

Facet B--'belief' is weaker than 'experience' (overt
behavior) in being 'passive' rather than

_ 'active.‘

Facet C--referring to the behavior of one's 'self'
(mine/my) rather than that of 'others' is
stronger in that it implies personal involve-
ment.

Facet D--'comparative' behavior is weaker than 'inter-
active' behavior. It does not imply social
contact, and a comparison is more passive than
interaction.

Facet E—-'hypothetical' behavior is weaker than
'operational.’ It does not imply acting out
behavior.

 

 

 

 

The above analysis is restricted to the ordering
implied in the five facets of Table l6--what Guttman is now

calling joint struction. However, an additional question

 

can be asked-—Is it possible to establish an ordering

175

principle so that the inter-item content itself can be
structioned or "ordered" with some explicit a priori semantic

meaning: i.e., rather than attempting to a posteriori

 

evolve the meaning by some procedure such as factor analysis?
Rokeach (1968) has independently developed and
made explicit the idea implied in the Jordan-Guttman paradigm
of Figure 1--the facet "y" of "condition" in Figure l is
equivalent to Rokeach's "situation;" one could also argue
that the entire lateral dimension of Figure l (facets F-J)
is equivalent to Rokeach's "object-specificity."
The rationale used in the selection of the item

content of the ABS—MR (and of the other Attitude Behavior

 

Scales that followed) attempted to "order" the item content
via three principles:

1. Ego involvement: Cognitive-affective. Is the
"attitude object in situation y" dealt with
cognitively or affectively?

2. Social distance: Distant-close. Is the "attitude
object in situation y" distant or close to one's
self?

3. Relevance: Low-high. Is "situation y" relevant
and/or important to the subject?

Consistent with the above discussion of the weak-
strong principle developed in Table 22, a positive or
stronger attitude would be expressed by a subject who
"agreed with or chose" items that dealt with the attitude
object in "highly important situations that involved the
self in close interpersonal action."

Two types of data analysis are indicated: (a) an

analysis of the facets across the six Levels, i.e., whether

176

or not the simplex is obtained, and (b) an analysis of the
scalar nature of the content within each of the six subscales.
The first analysis deals with the joint dimension and the
second with the lateral dimension (Jordan, 1968). The section
in this chapter titled "The Simplex Approximation and the
ABS" contains simplex data for the several different scales
thus far develOped by other researchers using the scale.

Joint struction refers to the difference between
subscales or Levels, or facets A through E of Figures 1-4.
Six additional facets, F through K, were added to differentiate
item content within Levels. These additional facets denote

item content and are labeled lateral struction. The complete

 

 

mapping sentence for the family of scales constructed, or

to be constructed, on this a priori basis is given in Figure 1.
Every item on every Level of a form of the ABS corresponds

to a combination of elements of each and every facet A through
K. The ordering system for lateral struction, however, has
not been developed as fully as has the system for joint
struction.

Jordan's (1970) research with the ABS-MR focused on
five questions or purposes: "(a) that the ABS-MR attitude
levels exhibit a simplex structure, (b) that relevant object-
situations were selected, (c) that selected variables are
effective predictors of favorable attitudes, (d) that the
ABS-MR can differentiate between groups, and (e) that the
ABS-MR is acceptable cross-culturally equivalent and com-

parable" (pp. 50-51).

.177

.coHuosuum HmHmHMH muocmp n QmsouQu m mumommo

.QOHuosuum HQHom muocwp m QmsouQu H muwomm

 

 

   

 

 

mHaH .emuo scum:

mon mumou .mm0h cmm
Quummmmm HmcoHHMUSpm co
umcHEmm cmoHumEmuwucH umuHm

 

 

 

 

Q
.0 umomw w
EOQH mucmuon K mHHuHmom Mm .ouw H
wQu Ho mmcmm oQu . . . . . .; . m
CH QoHQmmHoHuHmm / NH cwrob H
. . . . H3 H
pumsou mocmHm> Hmuuswc . Qu. mquuo NH 0H
m>Hpnomc Hh HHmm HH
wmcmm mucmeHomm
,.
Any QHV
n.
HMQOHqumuucoc mm Op aHHmsuom (H
CH mmcho mm CH coHuamHoHQHmm 0Q wHQm 0Q NH ob [k j(mQOHfiBQQO mm
H MQ \ H 0H m omd Hm
HMUHmOHOQQomQ m copumxmu oQ on quso w Q Q» Q .
coHummHoHQHmm QoHummHoHuumm m .Qouom mo chEOQ
Hoe 111 Que
1 N
AQoHLQOQ HHmm m
HOUUG A%E|®CH§HV AMUMHO 3...)... H,o))) s .
mQu ’ HHmm m / oocmHHwam NQ \mn; (1; , (.a44meQmewmxm
). s .1 :xucsoo :L CH
:2: A, mama Q. 5.92: HEM. 7 a H...
’ mHuuman/ u . Q 1 Q
quom \ H0H>QOm ucmuwmmm / ucoumwom

HOV Amy

Adv

coHumoscm HuHmuw>Hcs

. 1‘] -—

"mQu um pmmH>mmm

HmcmEO3 .m.mv mmHou Q
Ammsupv
onuomum HmHoom

Quos

coHumospw Humpcoomm

coHumospw mumEHum

HNMVLOW

QoHumospm HmEHoHco:

 

Q
Q
Q
Q
Q
Q
m

mmu< coHummHoHunm

Amy

 

I QQHB muomumucH mp \
mmummfiou Hp .

‘ uOHmewm ’
msoumuwucH m.uouo¢ _

..

ADV

ucmumwmu ou

mmHSQHHuum va pothsm

.mmHuH>Hqu pmHHHommm 2H QOHummHUHuumm pum3oe muoH>Mme

ImpsuHQHQ mo mconcmEHo Hmumumq cam HQHOh mo mHmmHmcm umomm mQu mom

m

mocwucmm mchmdzun.H musmHm

178

 

nuoH .H 0:35
iHc: oH-un cauHQoHl
toHumuavu Ho ouoHHoo

cavuon .w econ

 

 

 

.oootauou HHHoucoe on» Qua: xnou uHonu cH .03-nH Lon-l

unu undue-«Lao; .uouou 0.0:» avg» con-u. cu-HuoHocoaon oaocu an nonsaoa on can an ruaoucu <1 nuouau on» go uneasy-«Lacuna.

on» go HquHHa> ucoucoo one .swuH .ucHLnn “huHoLo>H:: ouaum cauchux an LacHIo. ouanuatu - cH out: 0:) ac. oocunuou hHchcol
on» zuH: AHHL-IHLQ ocHxLox .90: on) .uaHooHoQoaoa Hoono- ucaoHuooun hate-u cuH: coHuouonoOu cH oo>Ho>o ago: .uouau uchH

 

 

 

 

 

 

 

 

 

 

 

. cuouoo a.
thou-o u. nae-u. H-Hoo. .4
vouasvoa can! no no>HuacLouHa m2» .naocucoua own-Q0 mx~ coHuausoo n4
vcHx «o ucvcHonco m: huHuoHuHHou wt ascocH ma
acacia Ho “caoca Hz. . onzuoauu. oaHa> Ht. ~ on. H4.
uncanny o voHyocz uc- unnatuu HmqudHozoaaaanHuom uHcmatuoaoo
How Hz. HIV Hay
an vouoHooquuno neoatmm
novtnuot HQH-ucoe uu-no» nooauHuuc onus» on poo-H Quqo cflnpnx econ HHHLoHua 29H: out-Ho» oHu-Qo>aecs map
:oH m: Hatunoc wn
can ax oHn¢Lo>¢u «a.
quuoHnm oocoHn>
Hz. an.
vH- Hatouoc no
acuEQOHo>on HacoHuac aH ’ oHa eunu- .0
4|: wcH>ac LoH>aQoQ HaHoon nH Ho Head on» erase» Ho onoaraa on» to» noogzo. vHa HauoH no
“coeeono>~v HacoHHauo> «H “colouco>va ax ~ Houcotoa «o
“coEoucn>o¢ uHoc HH coHuuoqua ”x gnu. Ho
udddu onomunm LoHHaaam
AH. H11 Hue
H-onos Ha
Hcouaonouv voHeHuoanc: ha
voHvau- we HQLJuHaoxuchou on
oouHHcHuo- no ass-Lu H¢u¢:-unoa mm
voHcHooa-cs nu nouco>oua .n and noum.nane nooaau HauHcoucoo :0
Han. an , uteruo Hmfl. ’ c-cu egos nu vuuaHooH no HLJHcH Hence no
aflactuc mm , on 05-. ~m voaouuuo we ’vu vuaoru mo HancoEAQHo>oo No
a Cue» mm H ~ can» onoH Hm . nouooaoo Hn g yum HUM HuHuocunv atouanouoc Hm
.MHMHWI .
o>HHatamsoo “:05H-oth o>Huouama0n “coconuth tuHumnHa> nonaao
Hue Que Hue Hue Hm.

voutauuu AHH-ucolucoc ~<

 

 

 

 

..HsHsaaauuu HaoHnsza w< . Hones» w»
atoHootn HHHHccOHLoa m« ”Hnouniuno m»
Ion» uncosunnnvanc naHuo- a< ’ .aou a»
coHuaanHoL oHHE m< , “cu” m» ’ “you. an ~
:oHu-ut-uot ouatoooe n «Lu onxxuudx «canton ~ vaunwnoauan Hmnnhuon m» ’ ya; ”you to osofluma on wquQOHua ~x . H
coHu-vtauon uL~>aa H< couture» nnmnnu s H» ~ muons: Hx ’
neHucHLuuuchmmm pronouoa HHMNMMM

H<Q A: t.,?
.mcomumm pmcumumm

mHHmucmz pumsoe mopsuHuufi mo mHthmcd Humomm mQu How mucousmm mchmmz ¢I|.m musmHm

179

’ QwHQ MHfl . m>HuHmoa mx ~ HauoH>quQuHuoHnHQq MH
.auHmcoucH EsHume NH Hana mocon>V Hmpuamc mg QAH: muHHHQmme Hmsuom Ha o>HuooHuu HH
Q 30H Hv .o>Humwwc x Aflllll amoHuch pmuanHuupo w . o>HuHcm00 H

 

auHmcmucH mocmHm> Hm>mQ uHer omxa uHapa
Age Age 1H1 HHS

ZOHBUDmbm mooz mmzommum

acoHumHmQ asouw HLmUQOoum on
H HHHEmH uca xoa
:oHuaHmemH
cam QOHHmz .QuHaoQ mu AcH umnpmuua
acmeonan mu - aHHuucos
coHuaozum mg mg» Qquv

 

 

 
   

  

 

 

 

 

 

 

 

 

 

 

m m
mqumwp c . 30H m coHumeomp H
HnHmqu Ho mmsmomn mg HmommummmsH EanmE mm chH>Ho>cH ncoHumHmp macaw szEHQa mu
0» pomamwp Qsz HQ H H Hv cwHQ Hm mpHmQu Hmcomuoq Hg
unmuopm :oHumsHa>m mocmupomEH mcoHumsuHm ucHQ
Hxv HOV .Hmv
ZOHBUDmBm H<mm944
Axes HuoH>ezmn
N . N Heuom'bg N uuw>o N A V N
IMHHmcoHumpumd Ho Hmv pomtm c He as» t m 0.:50mmzu mocmHQm xv Ho mzrons HHom a How nouanuuum
ameOHquuoaNQ a any mam 506 c uaeuflmtmcpo Ho m:>-?aH> cmHHmQ Q so no Ha uoonnamv
QoH>onm n_u0uo< QoH>onm
co cHano msouwnmucH m.p0po< Houo< poH>mcmm acupuuom acopouom
Ame Hes Hoe Ame H<c

onboDmEm HZHOH

 

.coHQMpumumm HmucmzumHmom uoH>MQmm munuHuud 0Q» unauuauum on
com: mumomm coHuosuum woos wmcommmu cam .ucHOHme .ucHomcou mQu mo wocwucmm mchmmE Q .n .mHm

180

:.x= Qo30uQu =m: mquMH mm>Ho>cH coHuosuQm Hmumumqo

=.m: QmsouQu :<: muoomH mo>Ho>cH coHuosuum ucHonQ .cmpu0h paw mEmumEmm >Q pm>Ho>mm

I m>Hunom Nx \

 

 

 

 

 

 

 

 

 

 

 

 

mc0muod muHQz so QomHm pumzou mocmHm> w>Hummmc HQ ’ .fliQuHB
\ mocme>
va
mcHdoo HonH>MQmQ m
HmucoesuumcH mH I ouHmmwp mQ QmHQ m
I uHmuu Hanuum NH ‘ w>HQooHHm NH QHmQu Ho omsmooQ NQ \ coHumsHm>o ’ tszwE No ‘
v uHmQu pmusQHHuum Hfl o>HuHcoou HH \ o» Hoodmmu QuHB HQ ’. 30H Hm
Hm>oQ uHer v wowe uHmHB mmwoon coHumsHm>m \ wocmuuomEH ’
Hhv AH. Amy HOV
NHMQHHHE pcm ups NH
EmH>HQomnHmoHQHH0d wm
popuo pcm BmH mu
. maoH 4H
mchSOQ mu
N) NmecoHumuomo Nm N
m . coHumospm u 1- . . . Qqu muumumucH p
cH>Ho>cH . c. mtOmeu .
moHumHquomquo Hm muHcQ Ho zomHm (‘quSO\UH:m3 H H
. . ,.. _ >HHmuHu0QuooHQ m mmummEoo U
mcoHumsuHm QHHQ HoH>onm m_uobom uoH>onm
in Ho chEoo msowmwmucH m.u0uu«
Hmv Hay
kWV. m
chum Amt wcHeV N no on w QHoH>onQ upmyov mMQ «How m
mQu HHom o \ me m mmw ’ mocuHQonxo NQ ‘ OQB .uuo .poocmHHomxo ucmummmu Cu
QMQu H . . . .Auucsoo .HE CH 4 mwuanHuum CS womhnflm
\ muwQuo o ’ \ moHHoQ HQ mHoQuo Hm
HOuOd uoH>onm ucoummmm QQQHOHmm
HOV Ame Adv

.HmQuo Qomm pumsoa mmpzuHuu4 .mmuHQ3 cam .mHUMHm

ucHon Ho mHthmsd uwomm mQu How moqmucmm mchmmz|u.q mHSmHm

0 end can mum m cm
H u um OH H H p Q m

181

Jordan's research on attitude-behaviors toward the
mentally retarded has implications in three areas of concern:
(a) methodology, (b) determinants, origins or predictors of
attitude-behaviors, and (c) implications for attitude-
behavior change research. Jordan (1972) believes that facet
theory can be used in defining a problem; structuring rela-
tionships within and between variables; dealing with the
problems of relevancy, equivalency, and comparability; and
assisting in the analysis and interpretation of empirical
data. Other attributes and findings of Jordan's research
are that certain aspects of attitude-behavior are cross-
culturally invariate, i.e., the simplex is determined
largely by the structure of the object-subject relationship,
'certain' aspects of attitude-behavior are object specific,
situation specific, and/or culture specific, and that atti-
tude change must be approached multidimensionally: knowledge
being more related to Stereotypic and Normative Levels and
contact, values, and enjoyment factors more related to Actual
Feeling and Action (Behavior) Levels.

The ABS-MR is the first of a family of scales being
develOped by Jordan (1968) using the model presented in
Table 17. Scales have already been developed toward such
diverse attitude objects as the blind, deaf, war disabled
in Vietnam, drug users, and racial-ethnic groups (Erb,

1969; Frechette, 1970; Hamersma, 1969; Harrelson, 1970;

Harrelson, Jordan & Horn, 1972; Jordan, 1970, 1971a, 1971b;

182

Jordan and Maierle, 1969; Kaple, 1971; Maierle, 1969;
Poulos, 1970; Vurdelja, 1970; Williams, 1970). The com-
pleted mapping sentence for the family of scales constructed,
or to be constructed, on this a priori facet theory basis
is given in Figure 1. The attitude object of interest (i.e.,
the mentally retarded in Figure 3, and racial groups in
Figure 4) is simply substituted for "specified" persons in
Figure 2.

Other Attitude Behavior Scales (Jordan, 1970, pp.
47-48) currently available are as follows:

1. ABS-BW/WN......Blacks toward whites and whites
toward Negroes in seven areas
(C) Characteristics (Personal)
(E) Education
(H) Housing
(J) Jobs
(L) Law and Order
(P) Political Activism (Racial)
(W) War and Military
(G) General (two items from each of the above
seven)

2. ABS-SAF......"Africans"/Whites (in South Africa)
(G) General, minus L, P, and W items of the
ABS-VW/WN

3. ABS-MP/PM......Moari/Pakeha (New Zealand)
(E) Education
(G) General

4. ABS-MI or EDP......Mentally Ill or Emotionally
Disturbed Persons

5. ABS-DR......Deaf

6. ABS-ABE.....Undereducated Adults (Adult Basic
Education)

7. ABS-CES.....Black vs. White Cooperative Extension
Service farm agents

183

8. ABS-BL......Blind
9. ABS-WD......War Disabled (in Viet Nam)
10. ABS-DU......Drug Users
11. ABS-ENV.....Environment
12. ABS-WOM.....Role of Womena
13. ABS-TEC.....Technica1 Educationa
l4. ABS-EDC.....Educationa1 Changea
An ideal, complete research project, as Guttman has
elsewhere suggested, would consist of observing a value for
each subject on each variant of facets F through K for
each of the six Levels A through E (see Figures 1-4).
Guttman has further suggested that any coherent theory
referring to empirical research can be expressed in a similar
mapping sentence (Figure 5). He further states.
Lack of theoretical clarity as to the specifications of
the facets of the mapping may be the situation that
often impedes the connection between abstract theory
and empirical work (Guttman, 1959, p. 323).
Clearly the ABS-MR scale falls short of the ideal, complete
research project suggested by Guttman. Nevertheless, it
represents one of the few such attitude scales constructed
on an a priori basis according to Guttman's facet theory.
The ABS used in the present research, measuring

attitudes of blacks toward whites and whites toward blacks

was constructed by Hamersma and Jordan (1969) and revised

 

aScales 12-14 were developed with support by the
Organization of American States at the First Interamerican
Seminar on Educational Research, San Jose, Costa Rica,
March 6-24, 1972.

184

. a . ‘
Figure 5.--A Mapping Sentence for Strategies of Theory Development.

a receiving

 

1
Investigator (x) uses a strategy for a2 implementing ideas for his
a3 evaluating
b1 definition cl definition
b2 specification 02 specification
construction of a . for a construct .
b3 rationale c3 rationale
b4 hypothesis c4 hypothesis
dl assertion
d2 deletion
of his theory (y) through treatment (33 substitution of aspect (z) of
by . .
d4 intenSion
d5 extension
tu 1’
‘el concep a ( constructs that were constructed by ‘fl others 1
)e2 empirical ( ‘lfz himself(
5 gl high
:» 92 medium level of strategic formalism.
93 low

 

A Condensation of the Sentence

 

INVESTIGATOR (x) / uses IDEA STRATEGY (a) / for his CONSTRUCTION OF (b) /
for CONSTRUCT (c) / of his THEORY (y) / through TREATMENT (d) / of ASPECT
(z) / of CONSTRUCTS (e) / of CONSTRUCTER (f) / ..__._._.. LEVEL (9) of

strategic FORMALISM.

 

vv

An Abstraction of the Sentence

 

XABCYDZEF IE: aoG

 

 

aFrom Guttman (1971).

185

by Dell Orto (1970). Attitudes were measured in seven
separate attitude content areas. An eighth scale (G-General)
contained two items from each of the above seven content
areas. The areas were chosen on the basis of a number of
sources. The Report of the National Advisory Commission on
Civil Disorders (1968), a survey of twenty major cities,
lists the most frequently and consistently cited grievances
of residents of ghetto communities. AS stated by the
Commission's report, these grievances remained consistent
throughout every major city that was surveyed. "As the
Commission stated, these grievances were linked in a major

way to the attitudes that blacks and whites hold in relation

 

to each other. They ranked the deepest grievances into
three levels of relative intensity and presented them as
follows" (Hamersma, 1969, p. 84):

First Level of Intensity

1. Police Practices
2. Unemployment and underemployment
3. Inadequate housing

Second Level of Intensity

4. Inadequate education

5. Poor recreation facilities and programs

6. Ineffectiveness of the political structure and
grievance mechanisms

Third Level of Intensity

7. Disrespectful white attitudes

8. Discriminatory administration of justice

9. Inadequacy of federal programs
10. Inadequacy of municipal services

11. Discriminatory consumer and credit practices
12. Inadequate welfare programs

186

To comprehend more fully what the Commission meant
by these categories, two of them ("police practices" and
"inadequate education") are described here as they were
explained in the Report (1968):

Police practices were, in some form, a significant
grievance in virtually all cities and were often one
of the most serious complaints. Included in this
category were complaints about physical or verbal
abuse of Negro citizens by police officers, the

lack of adequate channels for complaints against

the police, discriminatory police employment and
promotion practices, a general lack of respect for
Negroes by police officers, and the failure of police
departments to provide adequate protection for
Negroes.

 

The educational system was a source of grievance in
almost all the 20 cities and appeared to be one of
the most serious complaints in half of them. These
grievances centered on the prevalence of de facto
segregation, the poor quality of instruction and
facilities, deficiencies in the curriculum in the
public schools (particularly because no Negro history
was taught), inadequate representation of Negroes

on school boards, and the absence or inadequacy of
vocational training (p. 144).

 

In addition to these grievances cited by the Report,
Hamersma reviewed additional research (Brink and Harris,
1964; Brink and Harris, 1967; Shaw and Wright, 1967; CBS
News, 1968; and Maccoby and Funkhouser, 1968) in the area
of racial attitudes and racial discontent. These reviews
cited similar areas such as housing, personal characteris-
tics, law enforcement, and unemployment as crucial.

Hamersma used these sources, the Report's grievance levels,
and suggestions from personnel of the Urban Adult Education

Institute and the Foundation for Racial Equality: In

187

Memory of Martin Luther King, Jr. in Detroit,1 to construct
the scales dealing with seven attitude content areas:

(a) Characteristics-Personal, (b) Education, (c) Housing,
(d) Jobs, (e) Law and Order, (f) Political Activism-Racial,
(g) War and Military.

For each of the seven separate attitude content
areas, a six-Level scale was constructed in accordance with
Jordan's Six-Level adaptation (Table 6) of Guttman's original
four-Level paradigm for attitude item construction. Each
attitude content area scale contained the Six Levels of:

(a) Societal Stereotype,(b) Societal Norm, (c) Personal
Moral Evaluation, (d) Personal Hypothetical Action, (e)
Personal Feelings, and (f) Actual Personal Action, as Shown
in Table 6.

The difficulties in constructing items and building
an item pool from which to select items for an attitude
scale or survey has been succinctly put by Ostrom (1971-1972,
pp. 593-594).

Problems of item construction were recognized by

the earliest workers in attitude measurement. At
least five papers on this subject had appeared prior
to 1940. Unfortunately, the guidance offered by
these investigators did not extend beyond listing
grammatical and stylistic considerations; classifying

items into such categories as belief, judgment, atti-
tude, and fact; and recommending that all possible

 

1The Urban Adult Education Institute and the Foundation
are concerned with providing adult education to people,
mostly black, who have not completed school. They provided
assistance in several phases of Hamersma's study.

188

attitude stands be represented in the pool. Tech-
niques suggested for item collection by these sources
are limited to such informal suggestions as inter-
viewing respondents who are known to have differing
attitudes, reading relevant published sources, and
relying on one's own command of the issues involved.
Guidelines provided by more recent treatments of
attitude measurement do not provide any more advanced
instructions for item construction.

Ostrom (1971-1972, p. 594) states that "the central
problems in assemblying an item pool lie in defining a
universe of content and in insuring an adequate sampling

of that universe,"

The Simplex Approximation and the ABS

 

As previously stated in discussion of the Contiguity
Hypothesis, subuniverses Closer to each other in the semantic
scales of their definitions will be closer statistically.

Kaiser (1962) suggests a procedure for testing a
simplex approximation: "for scaling the variables of a
Guttman simplex . . . the procedure . . . orders the vari-
ables. A measure of goodness of fit of the scale to the
data is suggested" (p. 155).

Kaiser's (1962) approach may be seen as performing
two functions: (a) a "sorting" of virtually all possible
adjacent pairs of matrix entrees so as to generate the "best"
empirically possible simplex approximation; and (b) an
assessment of a descriptive statistic, with a range of 0.00
to 1.00.

A computer program was developed which (a) re-ordered

the adjacent pairs of level members of each matrix, by

189

Kaiser's (1962) procedure, so as to generate the best empiri-
cally possible simplex approximation; and (b) calculated
Q2 for the hypothesized ordering<xfand.for the empirically

best ordering of each matrix.

Appropriate likelihood ratios are not yet developed

 

to fully assess the simplex approximation. Mukherjee
(1966) suggests a method which appears appropriate for
matrices of equally spaced correlations but neither Maierle
(1969) and Jordan's (1968) facet theory nor the actual data
suggest that the matrices in the present study have equally
Spaced entities. Harrelson (1969) discusses the Q2 method
in detail.

Hamersma (1969) suggested that "6-reversals" should
be the maximum possible in a 6x6 data matrix to still
consider it as 'approximating' a Simplex. By the "6-
reversal" criteria a Q? value of .60 would appear minimal
and preferably a value of .70 for a 6x6 matrix to be
acceptable as a Simplex (Jordan, 1970). AS indicated in
Table 23 (matrix 12.4), the highest g2 value is for an

ordered matrix containing both equal-interval and largest

 

correlation values. Thus far, the results of the ABS

appear quite favorable. Many of the early studies using
the MR scale reached this level of simplex approximation,
and more recent modifications of the ABS have yielded g2

scores approaching 1.00.

190

.OOQHHHOOQS mum Hmpuo mo

mHmme>mmm

 

 

 

 

 

 

 

 

O m v m N H w m v m N H O m v m N H mHm>mH
II 00 Om OV mm ON O II om Om Oh om Om 0 II mm Nb hm MH NO 0
II 00 Om 0v mm. m II OO om on 00 m II mm mm ON no m
II 00 Om Om v II om cm on v II om bv mN v :pmuwpuo:
I: oo Om m I: Om om m :1 mm mm m mum
I: nu 1: mm H» m
e.mH mm m e.mH mm M m.~H mm m o. u z
xHHumS xHHumE xHHumE H
mom. u maummm Ham. n ma ummm wow. u maummm
:HMMBBH :ERRHEH :EEHBEH
--ONBHR w --EHRE a --smmz H
II OH» Om 00 ...I oo om om m. II no mo hm m.
.II.II .II.II II.|I. m muo
ll. 0m 0? a. II om ON; MW. .ll mm ON T. :mu HQE m:
-- H. p -- am p .. mm m 93
m.NH In m.NH II H.NH mOOHHumz
xHHumE xHHumE xHHme I:
owe. n ma HmchHuo Hem. n ma HmeHmHuo 04. n ma HmeHmHuo
EHOB
mmosmnmmmHo Hmswm mwosmHOHMHo Hmsqm mmOQOHOHHHO Hmawmsa m>HumHHommO

 

.mcoHumeuuou smmsumm moosmummmHo Hmsqmcp

tam Hmsqm Qqu mmOHuumz :pmnwpuo NHHOOHHOH: HO MchHHMHmHHOO HMOHUOHOOQB mo mHmNHmc¢nl.mN mqmde

191

Early results using the ABS-MR are illustrated in
Table 24. The ABS-MR was administered to three groups in
the test develOpment study: (a) 88 Michigan State University
(MSU) graduate students (46 female, 42 male) in a course on
medical information for special education or rehabilitation
counselors-SER: students who were studying to be profession-
als in the area of disabling or handicapping conditions,
(b) ED 200—633 regular education students during the 1968
winter term, and (c) 523 elementary school teachers (381
female, 142 male) in Belize. All three groups yield the
simplex approximation pattern. The Q2 values for the SER
sample were the same (.97) for the hypothesized order and
the best order. For the ED 200 sample, the Q2 values were
also the same (.94) for the hypothesized order and the
best order. The 92 values for the Belize sample were (a)
hypothesized order (.858) and (b) best order (.859)
Table 24). All these values exceed Hamersma's minimal
criteria g2 of .70 for a 6x6 matrix to be acceptable as a
simplex.

The simplex relationship has also been obtained for
other, more recent studies, using the ABS-MR (Gottlieb, 1973;
Harrelson, 1970; Harrelson, Jordan, & Horn, 1972; Jordan, &
Horn, 1972; Jordan, 1970, 1971b; Morin, 1969; and Vurdelja,
1970). Studies using modifications of the ABS-MR have
resulted in simplex approximation. Poulos (1970) develOped

an Attitude Behavior Scale measuring attitudes toward the

192

.Hom .m .onHV amouoe soup

.coHumHuommp mHmEmm now uxwu mom

Q

.pmcHHHmpss mum mHmmHm>mmm

 

 

 

 

 

 

 

 

 

 

 

II Hm MH Ho mo Ho o II.mH Hm mH oo HH o II mm mH mo Ho Ho o
II em mm MH «o H II mm NH NH NH m II mm mH.NH HH m
II we NH oH a II mm mm Hm a II mm Hm NH H
II em HH m II mm HH m II Hm mm m mHo>mH Ho =umeuo: moo
o.mH II om m e.mH II mm m m.mH II sq m xHuume some ocHemxm
xHHumE II H xHHumE II H xHHumE II H
no. u mNumom om. n mNumom mm. n ma umom
II Hm mH Hm mo Ho o II oH mm mH oH HH o II mm aH mo Ho Ho o coHuoH Hmcomuoa
II em mm HH «o m II Hm mH mm HH m II om aH NH HH m ocHHmmm Hmcomuoa
II we NH oH a II mm mm Hm H II mm Hm mH e coHuoH .uonuoasm Hmcomuma
II em HH H II mm HH m II Hm mm m coHpmsHm>m Hmuoz Hmcomuma
m.MH II mm m m.MH II mm m H.MH II we m euoz HmuoHoom
xHHumE II H xHHumE II H xHHumE II H mmwuomumum HmumHOOm
Hm. I maHmchHuo om. I ma HmchHuo mm. u mNHmchHmm
o m e m m H o m e m m H o m H m m H
oHoemm we I mum oHoemm mmm I oNHHom mHEEHm mmo I oom om ehoa o>HuoHuomoo

 

.oom om oQu mow mumo

Q.mmHmEmm mmm can .mNHme
ucmEmoHo>mo umwe szme mQu mo mmcoHHMHOHHOO memEHm mo mHmHHMQQII.VN mqmde

193

deaf. His data yielded a simplex for all best order 92
groups (.83-.93) and four out of five hypothesized (original)
order g2 groups (.58-.90) (Table 25). Frechette's (1970)
study of attitudes of French and English speaking Canadians
toward West Indian Immigrants also yielded a simplex approxi-
mation pattern (Table 26); his hypothesized g2 scores ranged
from .54-.91, while the best order g2 matrix values ranged
from .76-.93. Williams (1970), using the ABS:BW/WN scale,
found hypothesized g2 scores (.73-.90) and best g2 scores
(.85-.95) (Table 27). As the ABS is revised, closer
approximations to the 'perfect' approximation simplex should
result.

A most recent develOpment in the family of Attitude

Behavior Scales has been the "drug scale" developed by

 

Jordan, Kaple, and Nicholson (Kaple, 1971). Their simplex
results have been the most successful thus far. Kaple's
(1971) study used further refinements in the ABS scale.
As seen in Table 28, his g2 not only exceeded all other ABS
results thus far, but approached 1.00, a theoretically
perfect simplex. Guttman (1954a) defines a perfect scale
(or simplex) as the following: ". . . by a perfect scale
we mean a set of items such that each item separately can
have its categories put into a one-to-one correspondence
with intervals of the same continuum" (p. 223).

For the case of a perfect scale of qualitative data,

there is but one elementary component-the underlying
rank order. From a person's scale rank, one can

194

 

.AHHHIHHH .ma .oanc moHamm scum ”muoz

 

 

 

 

 

 

 

 

 

 

II mH. «H. AH. om. oH. II mm. 9H. cm. II mm. mm. HH. II mm. mo. HH. 50. II we. om. HH. Hm. Ho. coHuuc .0
II vH. HH. Hm. om. II mm. mm. II om. HH. II Hm. wv. HH. II mH. mH. mo. vH. mcHHomm .m
II 3. mm. mm. II no. II Ho. II mm. 3. II oo. 3. Ho. HmuHuwnuom>x .v
II me. vH. II II II mm. II Hm. mH. .Ho>m Hugo: .H
II . . II . w>HuMEuoz .
om.vH xHuumz mv mH.qH stscz SH.HH xHuunz QH.QH stumz om mH.vH xHuunz oH . m
mm. u om II om. I On me. I Icm Hm. I om II om. I am II waHuoououm .H
m m p m m
II mH. «H. HH. Hm. om. II mm. mm. mm. II Hm. .m. Ho. II HH. mm. 90. so. II Hm. ow. Ho. HH. cm. coHuo< .0
II vH. mH. om. oH. II oH. mo. II Hm. HH. II aw. mm. mm. II mo. wH. Hm. mH. mcHwam .m
II 2... mm. mm. II om. II Hm. II Hm. HH. II 3. mm. 2. HmoHumcuom>x .v
II me. «H. II II II oH. II oH. Ho. .Hm>m Hauo: .H
II . . II I ®>HUMEO .
9H.vH xHuumz me mH.vH xHuucz HH.HH xHuuux HH.¢H xHuucz om HH.vH xHuumz cm I z m
mm. I oo II mm. u xx. um. I at cm. I ,xo II om. u oc II omsuoouwum .H
m m m m m
o H v H m H a m w H o H v H o m w H H o m. H H m H
.nsooIIzcz .eccoIIozz .nscoIIcax .nsooIIemm .nEOoIImos
II vH. oH. HH. qH. HH. II mm. am. so. II HH. no. Hm. II nH. vo. mH. Ho. II mH. no. mm. mo. co. coHuu< .0
II om. vH. mm. nH. II vH. He. II a. mo. II vs. HH. 50. II Hm. mm. He. Ho. mcHHmwm .m
II mH. Hm. Hm. II xH. II mH. II HH. vH. II Hm. oH. mo. HmoHuonuom>m .v
II vv. mH. II II II me. II mm. mH. .Hm>m Homo: .H
oH.HH xHuuaz II Hm. w.qH xruuaz m.HH xzzuc: H.HH xHuumx Hm. m.vH xHuunz II vH. w>Humsuoz .m
mm. H Om II n. H A. m. N ..c A. III Sm II D. n . ll 0Q.» Owuw .
m n, m m H. m m H. m e mom u um H
II VH. oH. HH. vH. HH. II no. em. so. II vm. mH. mo.I II Ho. HH. eo.I Ho. II mH. He. Ho. mm. Hm. :oHuu< .0
II om. vH. mm. mH. II vH. Ho. II so. mH.I II vs. an. Hm. II mo. oo.I mm. no. mcHmem .m
II mm. Hm. Hm. II mH. II Hm. II as. no. II vH. mH. oH. HmoHumzuonsz .v
II we. mH. II II II vH. II mH. mo. .Ho>m Hmuoz .H
«fig xCumz II Hm 5.3 .2322. m4; x732... méH xHuunz m... HéH xHuucE II hm. w>HquIHoz .N
am. u moo II we. I moo 3.. I who om. n moo II mm. H moo II wa>u0muwum .H
o m H H m H o H v H o m v H o H v H H o m e H m H
Ehwe w>HumHnommo
.UCOUII242 .uCOUIIoz: .uCOUIImQZ .ucouIIBmm .ucOUIIhOP

 

 

Inn! 0

 

.mclmmd wcu mo ch>wH me co Manchu coucowom new wustmm waQEHMII.mN mqm<9

1595

.HmH.o .oooHo muumnumum scum

 

 

 

 

 

 

 

.mm. H Hm>mH mo. um M mo 05Hm> HmoHuHuUw mH. u Hw>wH mo. um M.uo 09Hm> HMUHuHHUU
mm. u Hw>mH mo. an M mo msHm> HmoHuHuUm .COHuQHuomwo meemm How axon mme
me. u Hm>mH mo. um.m mo wsHm> HmoHuHuuo .UwcHHumccn who mHmmum>wmm
II mm HI. om mo Ho II HH Hm Ho mo HH II oo oH HH. oH Ho II om NH om mH mm coHuua Hucomumo
II HH. «H. mm Ho II Ho Ho HH Ho II HH oH HH om II oH Hv mH Ho mcHHmmm Hocomumm
II HH mm. om II mm .mm .Mm II mm «H NH II no He mo coHuoa HHUHumsuomHm Hmcomumm
II mm mm. II mm mo II MN Hm II mm mH coHumsHm>m Homo: HMCOmHmm
II oH II Hm II Hv II Hm suoz HmumHoom
m.mH o.mH H.HH m.mH
xHuumE II xHuumE II xHHumE II xHHumE II wmxuowumuw HmuwHoom
. umwm . n I umwm . n m umm . mm
mm m om m Ho mt m mm m u m
II mm so mo mm oH II HH Hm Ho oo HH II mH HH oH Hm HH II om om HH mH mo coHuua Hmcomumo
II HH om NH Hm II Ho Ho HH Ho II oH oH HH Ho II Hv oH HH .Hm ocHHmmo Hmcomumo
II Ho NH om II mm mm .Mm II oo oH mm II no oH mH coHuoa HmoHumsuomHm Hmcomumo
II mm Hm II oH oo II HH Hm II Hq mu coHumsHm>m Hmuoz Hmcomumo
II flm II om II Mfl II mm Euoz HmumHoom
H.mH H.HH H.mH H.HH
xHuumE II xHHumE II xHuumE II xHuucE II mm>uowumum HMumHoom
Hm. n ma HmcHoHuo om. n ma HmcHoHuo mo. I ma HmcHoHuo Hm. I mo HmcHoHuo
o m H H m H o m H H m H o m H H m H o m w H m H
omHQEHm mm I 2H mmHQEHm Hm I oH omHQEHm om mIemm umHasmm om I om
.mmHmEdm wnu HON 0I23\3m Immfi mcu mo mcoHumeuuou wamEHm mo mHm>Hmc<II.oN mqmde

Q

m

196

 

 

 

 

 

 

 

 

.Hoo .m .oHoHV memHHHHz scum "muoz
II oH mm Ho HH om II oH oH HH mm Ho II om mH Ho HH oH II vH Ho mo so Ho 0
II oH mm mH mo II mm mm HH Ho II oH mm oo mo II mm mH oH mH H
II Hm HH so II mm mm om II ow HH HH II ov HH «H H
II om mH II mm mm II HH oH II ow mm H
II om II Hm II HH II HH m
m.oH xHHumz II o.oH xHuumz II v.oH xHHumz II m.oH xHHHmz II H
mo.umo ummm mo.nmo ummm mo.umo ummm mo.umo ummm
II mm oH Hm HH Ho II oH oH Ho HH mm II om Ho HH HH oH II Ho HH ow mm oH H
II oH Ho HH om II mm om mm om II mm oH oo mo II so mo HH Ho H
II mm oH Ho II Ho mH HH II ow HH oH II ow oH «H H
II om mH II mm mm II HH oH II HH om H
II mm II om II Hv II HH m
H.oH xHHpmz II m.oH xHHpmz II H.oH xHuumz II H.oH xHuumz II H
Ho.umo HmaHoHHo Hm.umo HmcHoHHo oo.umo HmcHoHuo Hm.umo HmaHoHuo
o m H H m H o H v H m H o H v H m H o m H H m H
o: om o3 mm
.mmsouo goummmmm map New zB\3mumm4 wnu mo mcoHuMHmuuoo memEHm mo mHmemchI.hm mqmda

197

. . 2 . .
TABLE 28.--Correlation Matrices and 2. Values for Original and Best

Simplex Approximations on the ABS-DU, All Categories,
Initial Scale.a

 

Original Simplex Matrix Total All Categories All Groups

 

__ Q? = .98

.55 -- matrix 17.1

.39 .27 --

.27 .25 .70 --

.24 .24 .62 .86 --

.21 .24 .59 .82 .88 --

 

Best Simplex Matrix Total All Categories All Groups

 

-- g2 = .98

.55 --

.39 .37 -- matrix 17.2

.27 .25 .70 --

.24 .24 .62 .86 --

.21 .24 .59 .82 .88 --

 

aFrom Kaple (1971, p. 102).

198

deduce his behavior on each and every item in the

universe of content being studied. Thus, a single

variable-the scale ranks-(or their transformation

into any other metric)-accounts for the totality

of behavior, and in particular for the inter-

relationships between the items (Guttman, 1954b,

p. 259).
Kaple's (1971) data resulted in a g? of .98; only one group
failed to exceed the .70 minimal requirement postulated by
Hamersma (1969) for an acceptable simplex approximation. A
set of tests whose observed intercorrelations form a perfect
simplex have a simple order of complexity (Guttman, 1954b,
p. 271).

It can be concluded that Kaple's modification of

Jordan's Attitude Behavior Scale, to test attitudes towards

 

drug use (ABS-DU), does indeed order attitude-behavior
along the six-Level continuum derived by Jordan, according
to a simple order of complexity. Kaple does not explain
his one 'reversal,‘ except to point out that the particular
group had an extremely low rate in questionnaire return.

He offers an explanation for his extremely good results

(Q2 = .98, the highest Q2 thus far obtained in the develop-
ment and modifidation of the ABS scales): "the directions
and item stems for Levels 3 and 6 were changed slightly
from previous ABS scales and the consistently high Q2
obtained may partially reflect the new semantics employed
at these levels." Kaple recommends that additional experi-
mentation and research is needed to further evaluate the
effect of slight differences in semantic structure (Kaple,

1971, p. 135).

199

Thus, the totality of attitude-behavior encompassing
the realm of attitudes and behavior as defined along McGuire's
(1969) entire cognitive-affectiVe-conative dimension and
ordering of this attitude-behavior from least complex
(cognitive) to most complex (conative), in the area of
attitudes toward drug users, has been analyzed according to
Guttman's 1959 facet design. The goal of this entire
research venture, begun ten years ago by Jordan at Michigan
State University, that of defining and facetizing attitude-
behavior as a necessary first step to producing means and
methods to effectively change attitude-behavior, has resulted

in the attainment of a "near-perfect" simplex.

Reliabilityyand Validity of the ABS Scales

 

Standard reliability procedures were applied to the

Attitude Behavior Scales since they were new scales. The

 

following section reports reliability and validity infor-

mation for both the original Mental Retardation (MR) scale

 

and the Racial Attitude (BW—WN) scale, a more recent scale

 

based on the Mental Retardation scale. An item analysis was

 

run on the ABS inter-item correlation patterns and item-to-
subscale correlations. The reliability coefficient correla-
tion for the ABS-MR and the ABS-BW/WN ranged between .70

and .95 (Jordan, 1971a; Morin, 1969). The method used for
reliability was the Hoyt (1967) method which produces a
coefficient similar to the Kuder-Richardson 20 measure of

internal consistency. The reliability coefficients found in

200

the studies compare favorably to those of many tests des-
cribed by Anastasi (1968) and with the bulk of other atti-
tude scales described by Shaw and Wright (1967).

"The content validity of the ABS-MR may be assumed
since the content of the items was evolved in cooperation
with practicing school psychologists in the field of mental
retardation and 'known' groups were used to assess predic-
tive ability" (Jordan, 1971a, p. 20). Facet theory guided
the selection of items and thus helped insure that the
item universe was sampled (Jordan, 1970, p. 33). Construction
of the attitude behavior scales in general, was based on
facet theory,and construction of the items followed a
systematic a priori method instead of by the method of
intuition or by the use of judges. Every item on every
Level of a form of the ABS corresponds to a combination of
elements of each and every facet.

Anastasi (1968) states that "content validity in-

volves essentially the systematic examination of the test
content to determine whether it covers a representative
sample of the behavior domain to be measured"(p. 100).
For content validity to exist, it must be encompassed into
the test while it is being developed. Anastasi suggests a
thorough and systematic examination of relevant subjective
material, as well as consultation with experts knowledgable
in the specific area. These suggestions were included in

the development and writing of the ABS-BW/WN. The items

201

were taken from a diverse compilation of resources in the
areas of racial attitudes and race relations. Subject
material, such as the Report of the National Advisory
Commission on Civil Disorders, was utilized, as well as
additional research (Brink and Harris, 1964; Brink and
Harris, 1967, Shaw and Wright, 1967; CBS News, 1968; and
Maccoby and Funkhouser, 1968) in the area of racial atti-
tudes and racial discontent. Suggestions from personnel of
the Urban Adult Education Institute and the Foundation for
Racial Equality: In Memory of Martin Luther King, Jr. in
Detroit were also used in constructing the scales.

Systematic analysis of all major aspects considered
by the test items is necessary for proper content sampling
(Anastasi, 1968). In the ABS-BW/WN attitudes were measured
in seven separate content areas, plus an additional general
area; these areas were based on the above sources of infor-
mation. The items in each of these seven areas followed
Guttman structure and were placed on Jordan's five-facet,
six-Level continuum, thus insuring known sampling of profiles
of the item universe.

Jordan (1970) indicates adequate similarity between
most of the three sets (i.e., the three samples) of corre-
lations for the ABS-MR test develOpment data. An analysis
of results of the six ABS-MR Levels yields additional
support for construct validity, since the postulated seman-

tic structure (cognitive, affective, conative) and obtained

202

statistical structure (i.e., the simplex) are in essential
agreement. "The scale level 'ordering' of l<2<3>4>5>6 in
several nations is further evidence of construct validity and
cross-cultural invariance" (p. 33). Jordan (1971a), Morin
(1969) and Vurdelja (1970) using the ABS-MR obtained the
simplex ordering predicted by Guttman's contiguity hypothesis
in 19 out of 25 groups tested.

Concurrent validity of the ABS—MR may be inferred
from the fact that the older, more experienced, and knowledg-
able samples also scored more positively toward the mentally
retarded. Data were gathered for the samples on twenty-two
predictor variables which offered considerable 'correlational'
evidence of the validity of the ABS-MR content, in that
groups with known characteristics responded as expected
(Gottlieb, 1973; Harrelson, 1970; Harrelson, Jordan, and
Horn, 1972; Morin, 1969; and Vurdelja, 1970). Jordan (l972b)
reports evidence for concurrent validity in that three ABS-MR
studies (Jordan, 1971a; Morin, 1969; Vurdelja, 1970) have
differentiated groups with known diverse characteristics in
several languages.

Evidence for construct validity is seen in analysis
of the simplex approximation data. The semantic structure
and the obtained statistical structure are in agreement.

The simplex results shown in Tables 23 and 24 show the
close approximation to the simplex that was obtained with

the early ABS-MR scales. Also, other studies using the

203

ABS-MR (Gottlieb, 1973; Harrelson, 1970; Harrelson, Jordan,
and Horn, 1972; Jordan, 1970, 1971a, 1971b; Morin, 1969;
and Vurdelja, 1970) have obtained results that fit the
simplex pattern. Studies (Frechette, 1970; Poulos, 1970;
Williams, 1970) using modifications of the ABS-MR scale
applied to other attitude objects have also resulted in
simplex approximations (Tables 25, 26 and 27). Kaple's
(1971) data yield an even closer approximation to the
'perfect' simplex pattern, illustrating that modifications
of the original ABS-MR scale applied to other groups can

lend support to construct validity.

CHAPTER VI

DESIGN

Those who live in a cold climate and in EurOpe are

full of spirit, but wanting in intelligence and

skill. [They] keep their freedom but have no

political organization, and are incapable of

ruling over others (Aristotle, Politics, cited

in Kovel, 1970, p. 13).

Instrumentation
The design for this dissertation is based on Jordan's

extensive research at Michigan State University described
in detail in Chapter V. Jordan expanded and refined Gutt-

man's four-Level facet design into a six-Level design,

maintaining the Guttman simplex approximation pattern.

Procedure

 

The Attitude Behavior Scales: Whites toward Blacks/

 

Negroes (ABS-W/B-N) were administered to 254 white Michigan
State University students enrolled in Education 450 during
the Winter Term of 1972. These students were all education
majors who had just completed student teaching and had
returned to Michigan State University to take this course
to complete their program for certification as teachers.

The two Attitude Behavior Scales (Appendix B) that were

 

administered to this group differed only in the subject-object

204

205

referent. One-half of the group received scales with the
term 'Negro' in them; the other half received scales with
the term 'black' substituted for 'Negro.' The two scales
were indexed so that every other scale was an "ABS toward
blacks" or an "ABS toward Negroes." Both scales were
similar in all other respects.

For both ease of sampling and application of the
scales later in the course for instructional purposes,
black and other non-white students were also given the

scale. Black students were given the Attitude Behavior

 

Scale that assessed black attitudes toward white people.
As this study is concerned only with white attitudes
toward black people, the non-white scale results were not
included in the data. However, the black students taking

the Attitude Behavior Scale Spring Term were given their

 

own results along with the white students in the class.
All data were anonymous as only the individual student
knew his subject number.

The data from this study were used for instruc-
tional purposes in two three-week sections of Education 450
offered Spring Term at Michigan State University. These
two sections were two of 74 'modules' offered to students
enrolled in the course. Of the 74 sections offered, students
could choose the section they desired. Students voluntarily
chose the "Attitudes Toward Minority Groups" section. The

Attitude Behavior Scales were used as an introduction to

206

material on "Attitudes Toward Minority Groups," the title
for these two sections.

The two forms ('black' and 'Negro') of the racial
Attitude Behavior Scale were administered to these sections
of Education 450 Spring Term (22 white students present in
the morning section and 25 white students in the afternoon
section when the scales were administered). To aid in
discovering affective relevance of the terms 'black' and
'Negro,' a Semantic Differential was also administered to
the two groups along with the ABS. The Semantic Differential
evaluated five words: 'friend,' 'enemy,' 'black person,‘
'white person,‘ and 'Negro person' on a series of twenty
word differentials. The concepts 'black person' and 'Negro
person' were randomized, each concept appearing before the
other one-half the time. These concepts were rated in three
areas on 20 scales: 'evaluation' (10 terms), 'potency'

(5 terms), and 'activity' (5 terms), in accordance with
previous studies conducted in the area of race and the
Semantic Differential (Osgood, Suci, and Tannenbaum, 1957;
Williams, 1964, 1966; Lessing and Zagorin, 1972). The
words included for each concept were the following: kind-
cruel, clean-dirty, bad-good, honest-dishonest, unpleasant-
pleasant, worthless-valuable, fair-unfair, cowardly-brave,
friendly-unfriendly, unsuccessful-successful, (evaluation);
thick-thin, strong-weak, small-large, soft-hard, heavy-

light (potency); and cold-hot, active-passive, slow-fast,

207

sharp-dull, violent-moderate (activity). The Semantic Diff-
erential scales (Appendix C) were given after the ABS was
completed.

An analysis was performed comparing the results of
the three scales of the Semantic Differential ('evaluation,'
'potency,' and 'activity') with the results of the Attitude

Behavior Scale for the Spring Term Education 450 samples.

 

The 'evaluation,' 'potency,' and 'activity' scales of the
Semantic Differential were correlated with each of the six
separate ABS Levels (Societal Stereotype, Societal Norm,
Personal Moral Evaluation, Personal Hypothetical Action,
Personal Feeling, Personal Action). The aim of this par-
ticular group of comparisons was to see if the E ('evalua-
tion'), P ('potency'), and A ('activity') dimensions of

the Semantic Differential related to the six Levels of the
ABS. Researchers (Osgood, Suci, and Tannenbaum, 1957;
Williams, 1969) feel that the E factor is a measure of
attitude. Jordan's work (l97la, 1971b, l972a)involves an
attempt to analyze attitude into six Levels. The present
analysis seeks to find out whether the E factor of the
Semantic Differential measures an aspect of attitude similar
to that on several Levels of Jordan's attitude-behavior
scales. It is an attempt to discover if high scorers on the
E factor also score high on the feeling and action Levels
(Levels 4-6) of the ABS. According to Osgood, Suci, and

Tannenbaum (1957), there is a weaker relationship for the

208

A and P factors in regard to attitude than there is for E.
The present analysis compares these factors with the six ABS
Levels to find out the relationship of each of the three
Semantic Differential scales with the six ABS Levels.

Efficacy, in this study, was measured by a five-
question 'efficacy' scale used by Hamersma (1969). The
scale was derived from research conducted by Wolf (1967).
Efficacy, as used by Hamersma, "purports to measure atti—
tudes toward man's effectiveness in the face of his natural
environment" (p. 98). Some of the research behind the
variable of efficacy includes a discussion by Katz and
Gurin (1969). They state that efficacy is the one charac-
teristic that "most clearly differentiates both children
and adults of two races" (p. 364). Further, they feel it
is closely related to achievement. Rotter (cited in Katz
and Gurin, 1969) calls this variable 'fate control.’
"Individuals vary in the extent to which they feel they can
extract material and social benefits from the environment
through their own efforts. In its broadest meaning, the
construct refers to one's sense of efficacy, or power, and
readiness to accept personal responsibility for what happens
to him" (Rotter, cited in Katz and Gurin, p. 364).

Rotter, Seeman, and Liverant (1962) discussed
broad implications of individuals who score high or low on
a measure of internal versus external control of reinforce-

ment. They hypothesized that individuals who have a high

209

belief in external control of reinforcement (i.e., these
individuals would supposedly score low on 'efficacy')
would be more passive than those having a low belief in
external control. People who felt in control of their
surrounding environment, would more likely tend to actively
change their position in relation to their environment.
Rotter et. al. hypothesized that individuals who are in
the middle range of the variable would be likely to increase
their own personal satisfactions by understanding the environ-
ment, even though they would feel they could not change the
environment much. These people adjust by learning "the
rules of the game" and although quite conforming, gain
maximum satisfaction through their conformity.
People with perhaps a still greater belief in inter-
nal control may include those who believe in their
own potential to change the environment or the world
around them. They are not merely ambitious but
could be creative, non-conformists, or revolutionary.
Their revolutionary goals might be in the political
realm of ideas or the arts. Not all peOple at the
extreme of internal control could be so characterized
because many of them might be highly rigid, moralistic,
or immobilized by feelings of failure. However, it
is quite possible that the real innovators could be
drawn from that population which is relatively high
in a generalized belief in internal control of
reinforcement (Rotter, Seeman, and Liverant, 1962,
p. 476).

Rotter (1966) distinguished between a belief in inter-
nal control and external control. An individual who believes
in a high degree of internal control (high efficacy score)
is one who feels that his own behavior and characteristics

can determine events that happen to him. A person with a

belief in external control, feels more at the mercy of his

210

environment. He believes that forces around him are not
entirely at his control; circumstances are considered more
unpredictableauuifate or luck is given as a rationale for
much that occurs to him. He perceives the world as more
complex and therefore difficult to change, effect, or
modify. Hypothetically, a person who believes in external
control would score low on the efficacy variable. Four
studies reported by Rotter confirm this analysis of internal
and external control of one's environment. The importance
of this variable of efficacy to racial attitudes and atti-
tude change has been summarized by Rotter (1966, p. 24) as
follows:

The individual who perceives that he does not have

control over what happens to him may conform or

may go along with suggestions when he chooses to

and when he is given a conscious alternative. How-

ever, if such suggestion or attempts at manipulation

are not to his benefit or if he perceives them as

subtle attempts to influence him without his aware-

ness, he reacts resistively. The findings have

considerable significance for the general area of
persuasion and prOpaganda.

Research Hypotheses

H-l: Whites taking the "ABS toward blacks" will have
significantly more negative attitudes than whites
taking the "ABS toward Negroes."

Rationale--Williams (1966) found that whites
evaluated 'black person‘ more unfavorably than they rated
'Negro person' on a Semantic Differential.

Lessing and Zagorin (1972) found that whites who

scored low on a measure of "black power orientation" rated

211

the two concepts, 'black person' and 'Negro person,‘
differently, rating 'black person' lower, but not signifi-
cantly lower, than 'Negro person.‘ Whites high in "black
power orientation" rated the two concepts about equally,
rating 'black person,‘ however, slightly higher than 'Negro
person.’ The researchers concluded there was no longer any
difference between the two terms.

The present author felt that white college students
in Education at Michigan State University come from fairly
conservative backgrounds and will evaluate 'Negro person'
more positively than 'black person,‘ primarily because the
concept 'black person' will elicit connotations of black
militancy and will hence be more threatening than the
concept 'Negro person.‘

Instrumentation--There were two forms of the ABS,

 

"ABS toward blacks" and "ABS toward Negroes."

Analysis--Multivariate analysis of variance (MANOVA).

H-2: The attitude data from the sample will form a Guttman
simplex.

Rationale—-According to Guttman's (1959) Contiguity

 

Hypothesis, Levels that are closer together semantically will
be closer statistically. The Level-by-Level correlation
matrix will approximate a simplex, unless the items were
incorrectly written or inaccurately assigned to Levels.

Instrumentation--Correlations between the scores of

 

the six Levels of the ABS-WB/WN-G.

212

Analysis--Kaiser g2

H-3: There is a positive relationship between a high
efficacy score and positive attitudes on the ABS.

Rationale-—A high score on this variable indicates
a person who feels in control of his environment and, there-
fore, less threatened by it. It is postulated that high
scorers will have more positive racial attitudes.

Hamersma (1969) found a positive relationship
between scores on the efficacy variable and favorable atti—
tudes toward the opposite racial group. Dell Orto (1970)
found a significant relationship for whites between efficacy
and positive attitudes at Level 4 (Personal Hypothetical
Action) and Level 5 (Personal Action) of the ABS: BW/WN-G.
Allport and Kramer (1946, cited in Dell Orto, 1970) found
that "those who were non-efficacious had a jungle philosophy
of life--viewing the world as basically evil and dangerous--
and were generally prejudiced" (p. 20).

The efficacy scale "was designed to measure atti-
tudes toward man and his environment and attempts to deter-
mine the respondent's view of the relationship between man
and his environment" (Hamersma, 1969, p. 98). The usage
and function of this scale was outlined by Wolf (1967, p. 113):

The continuum underlying this scale ranged from a
View that man is at the mercy of his environment

and could only hope to secure some measure of
adjustment to forces outside of himself, to a view
that man could gain complete mastery of his physical

and social environment, and use it for his own
purpose.

213

Hamersma (1969, p. 98) termed this variable
'efficacy' because it "purports to measure attitudes toward
man's effectiveness in the face of his natural environment."

The efficacy scale was one of five attitude scales
constructed by Wolf (1967). This particular scale involved
testing students' views concerning man and his surrounding,
external environment. The scale attempts to measure a
continuum of man's attitudes, ranging from viewing man as
at the complete mercy of his own environment, only able to
obtain some minimal measure of control through science
over the forces that surround him, "to a view that man
could gain complete mastery of his physical and social
environment and use it for his own purposes" (p. 113).

The five scales went through several stages of
develOpment before they were field tested on persons in
seven countries, including the United States. In each
country, 150 students from each of several age levels
were involved in the testing. The scales were then sub-
jected to Guttman unidimensional scale analysis, at which
time the items were again evaluated, modified, and refined.
According to Wolf (1967, p. 118), "the coefficients of
reproducibility for the final scales generally ranged
above the .80 to .85 considered acceptable by other

researchers."

 

214

Instrumentation-—The Life Situations Scale by Wolf

 

(l967, p. 122) involving nine items. The nine items on the
Attitude Behavior Scale attempting to measure efficacy

include items 101-109.

Analysis--Correlation coefficients.

H-4: Subjects taking the Semantic Differential will give
the concept 'Negro person' a significantly higher
rating on the 'evaluation' dimension than the con-
cept 'black person.‘

Rationale--See H-l, above.

Instrumentation--Semantic Differential scores on the

 

'evaluation' dimension.

Analysis--Two-samp1e t-test

H-S: Subjects taking the Semantic Differential will give
the concept 'black person' a significantly higher
rating on the 'potency' dimension of the Semantic
Differential than the concept 'Negro person.‘

Rationale--Lessing and Zagorin (1972, p. 70) found

consistent results regarding the 'potency' and the 'activity'
factors of the Semantic Differential evaluated by both black
and white students:

Blacks were seen as the strongest, most potent and
most active. Negroes were viewed as slightly less
strong and active, whites as still less strong and
active, with colored persons being viewed as the
weakest and most passive. Though the word "Negro"
was once held to have the meaning of "slave" by a
pre-Civil War court (Isaacs, 1963, p. 65), the
contention of Baird (cited in Bennett, 1967, p. 52)
that "Negro" evokes a slave image in the minds of
twentieth century Americans is questionable on the
basis of the present findings: Negro person was
perceived as no weaker than white person. Clearly,
however, the designation "black person" was the
most evocative of the powerful image which black power
advocates desire for Afro-Americans.

215

 

Instrumentation--Correlations between 'black person'
and 'Negro person' on the 'potency' dimension of the Semantic
Differential.

Analysis--Two-sample t-test.

H-6: Subjects taking the Semantic Differential will give
the concept 'black person' a significantly higher
rating on the 'activity' dimension of the Semantic
Differential than the concept 'Negro person.‘

Rationale--See H-S, above.

 

Instrumentation--Ratings of 'black person' and 'Negro

 

person' on the 'activity' dimension of the Semantic Differ-
ential.

Analysis--Two-sample t-test.

H-7: A higher efficacy score on the ABS will be correlated
with more positive scores on the 'evaluation' dimen-
sion of the Semantic Differential for the concepts

'friend,' 'black person,‘ 'white person,‘ 'Negro
person,‘ and more negative scores for the concept
'enemy.’

Rationale-—Those persons who feel in control of the

 

environment (score high on the efficacy scale) should have

a clearer and more positive conception of what 'friend' means

to them. They will have more positive attitudes toward

both their own group ('white persons') and other groups

('black persons,‘ 'Negro persons'). It is also believed

that they will differentiate between groups they accept

and like and groups they do not, hence giving a more nega-

tive response to the concept 'enemy.‘
Instrumentation--Correlations between efficacy

scores on the ABS and scores on the concepts 'friend,'

216

'black person,‘ 'white person,‘ 'Negro person,‘ and 'enemy'
on the 'evaluation' dimension of the Semantic Differential.

Analysis--Correlation coefficients.

H-8: There will be a positive relationship between high
scores on the 'evaluation' dimension of the concept
'black person' and 'Negro person' of the Semantic
Differential and positive scores on the ABS.

Rationale--Osgood, Suci, and Tannenbaum (1957)

 

define attitude as predispositions to respond distinguish-
able from other states of readiness in that they predispose
toward an evaluative response. "Score variation along the
E dimension covaries closely with the score variation on
conventional attitude tests" (p. 193). They further state
that studies completed using Thurstone and Guttman scales
lend additional support to the notion that the E dimension
of the Semantic Differential can be considered a measure of
attitude.

Fishbein (1965) stated that Osgood equated the
'evaluative' dimension with attitude, a concept's favor-
ableness or unfavorableness, goodness or badness. Fishbein
stated that Osgood defined attitude as a "predisposition to
respond;" this appears contrary to Jordan's (1971a) defini-
tion of "attitude as behavior;" however, Osgood's definition
is similar to what Jordan is attempting to measure and
according to Fishbein, can be divided into six categories,
including cognitive, affective, and behavioral intentions

or action components. These six categories were as follows:

217

l. Beliefs about the component parts of the object;
2. Beliefs about the characteristics, qualities, or
attributes of the object;
3. Beliefs about the objects relations with other
objects or concepts;
4. Beliefs about whether the object will lead to or
block the attainment of various goals or “valued
states;"
5. Beliefs about what should be done with respect to
the object;
6. Beliefs about what the object should, or should not,
be allowed to do (Fishbein, 1965, pp. 110-111).
Osgood, Suci, and Tannenbaum (1957, p. 190) stated
the following in regard to attitude and the 'evaluative'
dimension of the Semantic Differential. "It seems reasonable
to identify attitude, as it is ordinarily conceived in both
lay and scientific language, with the evaluative dimension
of the total semantic space, as this is isolated in the
factorization of meaningful judgments." They did go on to
suggest that other dimensions combined with E can improve
prediction of attitudes, but build the strongest case for
the 'evaluative' factor, and secondarily for the 'potency'
and 'activity' factors.

"The evaluative meanings of both the skin color,
per 22! and the color code term appear to be significantly
related to attitudes toward this group of persons . . . the
designation of racial groups by color names is one deter-
minant of racial attitudes" (Williams, 1969, p. 385).

The ABS is an attempt to measure "attitude-behaviors"

along a cognitive-affective-conative analysis. This is

consistent with McGuire's (1969) analysis of attitude into

218

knowing, feeling, and acting. Jordan uses Guttman's defini-
tion of attitude as "a delimited totality of behavior with
respect to something" (1950b, p. 51). According to Jordan
(l971a), this definition of Guttman's "is consonant with a
structural (Foa, 1966, 1968; Foa and Turner, 1970) approach

to the facet analysis of attitude-behavior" (Jordan, 1971a,

p. 7). Jordan and his associates at Michigan State University
have conducted several studies (Bray, 1972, 1973; Dell Orto,
1970; Frechette, 1970; Williams, 1970) indicating that the
ABS/BW-WN is a measure of racial-ethnic attitudes.

The 'evaluation' factor of the Semantic Differential
purports to be a measure of attitude along the cognitive-
affective-conative analysis; the ABS also represents a
measure of attitude along the same trichotomy of attitude.
It is hypothesized that since these two scales propose to
measure the same thing, there should be a positive relation—
ship between scores on one and scores on the other, if
indeed they are both a measure of thinking, feeling, and
acting (i.e., 'attitude').

Instrumentation--Correlations between the ABS and

 

the 'evaluation' dimension of the Semantic Differential.

Analysis--Correlation coefficients.

H-9: There will be no significant correlations between
high scores on the 'potency' dimension of the con-
cepts 'black person' and 'Negro person' of the
Semantic Differential and positive attitudes on
the ABS.

219

Rationale--Studies (Osgood, Suci, and Tannenbaum,

 

1957; Williams, 1969, 1971) have shown that a high score on
the 'evaluation' factor is a measure of positive attitudes

whereas results concerning high scores on the 'potency' and
the 'activity' dimensions are equivocal. Osgood, Suci, and

Tannenbaum (1957) specifically defined attitude as a
learned implicit process which is potentially bipolar, varies
in its intensity, and mediates evaluative behavior" (p. 190).

Williams gt_gl., (1971, p. 222) stated that past
research has shown that the color name white has been rated
as good, weak, and active, while the color name black has
been rated as bad, strong, and passive. Research is con-
flicting with regard to 'potency' and 'activity.‘ Williams
(1966) found the most consistent data with the E dimension;
he also found it for P and A, but not as strong nor as

consistent.

Instrumentation--Correlations between scores on the

 

ABS and scores on the 'potency' factor of the Semantic
Differential.
Analysis--Correlation coefficients.

H-lO: There will be no significant correlations between
high scores on the 'activity' dimension of the
concepts 'black person' and 'Negro person' of the
Semantic Differential and positive attitudes on
the ABS.

220

Instrumentation--Correlations between scores on the

 

ABS and scores on the 'activity' dimension of the Semantic
Differential.

Analysis--Correlation coefficients.

CHAPTER VII

ANALYSIS OF THE DATA

The one thing to keep uppermost in our minds is that
whatever is good for the black community will enhance
the white community because these communities have
always been and always will be inseparable even
within their separateness (Sikes, 1971, p. 104).

This study investigated several hypotheses con-

cerned with two forms of the racial Attitude Behavior Scale--

 

ABS/WB-WN, and a form of the Semantic Differential. Two
forms of the ABS were used, testing white "attitudes toward
blacks" and white "attitudes toward Negroes." The Semantic
Differential was used to investigate white students' atti-
tude meanings toward the concepts 'friend,' 'enemy,' 'black,‘
'white,' and 'Negro' on the three dimensions of 'evaluation,'
'potency,' and 'activity.‘ The two scales, ABS and Semantic

Differential, were compared to one another.

Analysis Procedure

 

The data were analyzed on the CDC 3600 and CDC 6500
at the Michigan State University Computer Center. Table 29

contains the basic variables list used in this study.

221

TABLE 29.--ABS-BW/WN/WB:

222

. . a
BaSic Variables

List by IBM Card and Column.

 

 

 

 

 

 

 

 

 

 

 

Brodwin Study (U.S. = 133).
IBM ABS-BW/WN/WB
Variable Range
Card Column Page Item
1. Stereotype 14-42 1 21-34 1-2 1-14
m 2. Normative 14-42 2 21-34 3-4 15-28
r3 3. Moral eval. 14-42 3 21-34 5-6 29-42
H 4. Hypothetical 14-42 4 21-34 7-8 43-56
3 5. Feeling 14-42 5 21-34 9-10 57—70
a 6. Action 14—42 6 21-34 11-12 71-84
7. Sex 1-2 1-6 36 13 85
8. Age 1-5 1-6 37 13 86
6 9. Marital 1-5 1-6 38 13 87
g 10. Religion 1-5 1-6 39 14 88
ll. Educ. -amt. 1-5 1-6 40 14 89
Change 12. Child rearing 1-4 1-6 41 14 9O
13. Birth control 1-4 1-6 42 14 91
p 14. Kind 1-3 1-6 43 15 92
8 15. Amount 1-5 1-6 44 15 93
g 16. Avoid 1-5 1-6 45 15 94
8 l7. Gain 1-5 1—6 46 15 95
18. Enjoy 1-5 1-6 47 16 96
19. Prejudice-reduce 1-5 1-6 48 16 97
8 20. Racial attitude 1—5 1-6 49 16 98
g 21. Ethnicity 1-5 1-6 50 16 99
22. Urbanity 1-4 1-6 51 17 100
Value 23. Efficacy 9-36 1-6 52-60 18-19 101-109
:3 24. Friend8 7 7 19-38
a 25. Enemy 7 7 40-59
. 26. Black 7 7 61-80
E 27. White 8 8 19-38
In 28. Negro 8 8 40-59
29. Nation (133) 001-999 1-6 1-3 -- --
5‘ 30. Groupéinterest 01-99 1-6 4-5 -- --
jg 31. Group-admin.C 01-99 1—6 6-7 -- -—
g 32. Subject no. 001-999 1-6 8-10 -- --
g 33. Card no. 1-6 1-6 11 -- --
34. Attitude object 1—6 1-6 12 -- --

 

a0n the 112271 edition of the

ABS-WB/BW

bFemale = 1, male = 2.

CAdmin. group

1:8:00 a.m.--Spring, 1972
2:12:00 noon--Spring, 1972
3=Winter term, 1972

d
Attitude object (col. 12)--
l = blaCk; 2 = Negro.

eEach "concept" has 20 bipolar terms

as follows:
Evaluation--19,21,23,25,27,29,31,33,
35,37

Potency--20,24,28,32,36

223

Descriptive Statistics

 

Two Frequency Column Count Programs (Clark, 1964)
were used to compile the frequency distributions for every
item in the instruments used in this study. This procedure
was useful to insure prOper and accurate representation of
the data in the computer prior to running it in computational

programs.

Statistical Analysis

 

In the CDC STATROUT Program (Ruble, Paulson, &
Rafter, 1966), a great amount of data can be employed in one
analysis. Separate analyses can be done for the total group
andiknrany number of sub-groups or partitionings of the
data. For each specified group, e.g., total, "ABS toward

blacks," "ABS toward Negroes,‘ etc., a number of statistics
can be requested. Those used for each partitioning in this
research were the means and standard deviations for each
variable and the matrix of simple correlations between all
variables.

Two sample t-tests for dependent samples were used

in the analysis as well as the Finn Program (Finn, 1970) for

multivariate analysis of variance (MANOVA).

Simplex Approximation
Kaiser (1962) has suggested a procedure for testing
for a simplex approximation. Kaiser's approach may be seen

as performing two functions: (a) the 'sorting' and

224

rearranging of all possible arrangements of adjacent pairs,
and (b) the assignment of a statistic, g2 to the original
and rearranged matrices. The index g2 is a descriptive one,
with a range of 0.00 to 1.00.

A computer program has been developed at Michigan
State University which will (a) reorder the obtained level
member correlations of each ABS: WB-WN matrix by Kaiser'sl
procedure to generate the 'best' empirically possible simplex
approximation, and (b) will calculate the Q2 for both the

obtained and the empirically best ordering of each matrix.

Significance Level

 

The .05 level is proposed as constituting signifi-
cance beyond chance for correlational coefficients,multi-
variate analysis of variance, and two-sample t-tests in

the present research.

Research Hypotheses

 

H-l: Whites taking the "ABS toward blacks" will
have significantly more negative attitudes
than whites taking the "ABS toward Negroes."

The results of the present study, using subjects

taking the two forms of the ABS, 'black' and 'Negro,'found

 

lAs documented elsewhere by Jordan (Harrelson, Jordan,
& Horn, 1972). Guttman has pointed out that the Kaiser pro-
cedure is limited to a simplex of the form r-k = aj/ak (j<k)
and alternate methods of simplex analysis are being explored
by Jordan and Guttman.

225

no significant differences in attitudes between the two terms.
Also, no differences were found among the six Levels of the
ABS and the three groups (one,Winter Term; two, morning and
afternoon sections, Spring Term) that had taken the attitude
scales; the results were reported both simultaneously and
separately for the six Levels. A multivariate analysis of
variance (MANOVA) was used to test these results. Tables 30
and 31 contain the results for the two racial terms and the
three Administration Groups, respectively.

H-2: The attitude data from the samples will form

a Guttman simplex.

The simplex approximation hypothesis was tested by
use of the CDC STATROUT computer program at the Michigan
State University Computer Center to produce Level to Level
correlations for all groups and categories. The Level to
Level correlations were then subjected to Kaiser's (1962)
simplex approximation test described in Chapter V. The
obtained matrix was submitted to a procedure that 'evaluates'
the obtained correlation matrix, resulting in a Q2 value.
The program also rearranges adjacent pairs of coefficients
into the best possible simplex order and computes a 'best
approximation' Q2. Table 32 illustrates the simplex
approximation g2 values for the three groups. For Groups
1 and 2, the g2 tables represent a combination of the two
forms of the ABS; this procedure was taken because of the

smaller samples for Groups 1 and 2. Since Group 3 was a

226

TABLE 30.--Mu1tivariate Analysis of Variance (MANOVA)
Results for Racial Terms, 'Black' and 'Negro'

 

 

Between

Variable Mean Sq. Univariate F P Less Than
Stereotypic 8.78 0.66 0.42
Normative 19.18 0.76 0.38
Moral Evaluation 10.98 0.21 0.64
Hypothetical

Action 6.08 0.16 0.69
Personal Feeling 15.18 0.49 0.48
Actual Action 76.64 0.83 0.36

 

Note: Degrees of Freedom for Hypothesis
Degrees of Freedom for Error
Test of Equality of Mean Vectors

291. P Less Than 0.6055

296.

0.7556.

1.

F--Ratio for Multivariate
= 6 and

TABLE 31.--Multivariate Analysis of Variance (MANOVA)
Results for the Three Administration Groups

 

 

 

Between

Variable Mean Sq. Univariate F P Less Than
Stereotypic 16.07 1.22 0.30
Normative 9.58 0.38 0.68
Moral Evaluation 114.58 2.23 0.11
Hypothetical

Action 85.93 2.22 0.11
Personal Feeling 17.50 0.57 0.57
Actual Action 311.79 3.36 0.04

Note: Degrees of Freedom for Hypothesis = 2.

Degrees of Freedom for Error = 2.96. F--Ratio for Multivariate
Test of Equality of Mean Vectors = 1.2938. D. F. = 12 and
582. P Less Than 0.2176.

227

. . 2 . .
TABLE 32.--Corre1ation Matrices and 9. Values for Original and Best
Simplex Approximations.

 

Original Simplex Matrix, Group 1

 

 

 

 

 

 

 

-- Q? = .86

.29 --

.06 .50 --

.12 .11 .46 —-

.16 .06 .39 .42 --

.05 .47 .15 .23 .22 --
Improved Simplex Matrix, Group 1

-- Q? = .86

.29 --

.06 .50 --

.12 .11 .46 --

.16 .06 .39 .42 --

.01 .47 .15 .23 .22 —-
Original Simplex Matrix, Group 2

_- {22 = .92

.23 --

.01 .12 --

.04 .25 .58 --

.05 .01 .68 .38 --

.03 .14 .48 .52 .21 --
Improved Simplex Matrix, Group 2

-_ g? = .93

.23 --

.04 .25 --

.03 .14 .58 --

.Ol .12 .68 .38 --
.05 .01 .48 .52 .21 --

228

TABLE 32.--Continued.

 

Original Simplex Matrix, Group 3, Attitude Object "Black"

 

__ g_ = .94

.47 —-

.13 .22 -~

.21 .13 .50 —-

.12 .05 .37 .60 —-

.07 .03 .04 .19 .17 --

 

Improved Simplex Matrix, Group 3, Attitude Object "Black"

 

-- 2

.47 -- Q, = .96

.22 .13 —-

.13 .21 .50 --

.05 .12 .37 .60 --

.03 .07 .04 .19 .17 --

 

Original Simplex Matrix, Group 3, Attitude Object "Negro"

 

-_ g? = .95

.37 —-

.26 .28 --

.09 .15 .33 --

.01 .11 .11 .31 --

.02 .04 .17 .28 .22 --

 

Improved Simplex Matrix, Group 3, Attitude Object "Negro"

 

__ g? = .95
.37 --

.26 .28 —-

.09 .15 .33 --

.02 .04 .17 .28 --

.01 .11 .11 .31 .22 --

229

TABLE 32.--Continued.

 

Original Simplex Matrix, Group 3, Attitude Object Combined

 

-- Q? = .95

.43 --

.18 .25 --

.15 .14 .41 --

.06 .02 .26 .46 --

.04 .03 .06 .23 .19 -—

 

Improved Simplex Matrix, Group 3, Attitude Object Combined

 

 

 

 

 

-- Q? = .96

.43 --

.25 .18 --

.14 .15 .41 --

.02 .06 .26 .46 --

.03 .04 .06 .23 .19 --
Original Simplex Matrix, All Groups Combined

-- Q? = .90

.41 --

.17 .26 --

.15 .15 .43 --

.07 .01 .26 .46 --

.05 .00 .07 .21 .17 —-
Improved Simplex Matrix, All Groups Combined

-- Q? = .93

.41 --

.26 .17 --

.15 .15 .43 --

.01 .07 .26 .46 --

.00 .05 .07 .21 .17 --

 

230

much larger group, the two forms of the ABS were analyzed
both separately and together. Table 32 presents the corre-
lation matrices and g2 values for both the original matrix
and for the 'best approximation' for every group and for
every category. Although Kaiser's (1962) simplex approxima-
tion test does not take into account the occurrence of nega-
tive correlations, few of the groups or categories had any
negative correlations in their simplexes.

Chapter V stated that a Q2 value of .70 is accept-
able as reflecting a satisfactory simplex approximation
according to the Jordan-Hamersma six-reversal criteria
(Hamersma, 1969). No correlation matrices failed to exceed
this criteria. The improved simplex matrices illustrated
in Table 32 were developed by a computer re-ordering of the
Levels to yield the best approximation to the simplex.

The data of Table 32 therefore, support this
hypothesis: the ABS: WB-WN does form an approximate
simplex. Close examination of the simplexes indicates
the possibility of reversal of 'order position' between
Levels 1 and 2. Further research could explore the feasi-
bility of either refinements in these Levels or the elimi-
nation of one of them.

H-3: There is a positive relationship between a

high efficacy score and positive attitudes
on the ABS.

The efficacy scale "was designed to measure atti-

tudes toward man and his environment and attempts to

231

determine the respondent's View of the relationship between
man and his environment" (Hamersma, 1969, p. 98). It was
postulated that persons who scored high on this variable
would have more positive attitudes as measured by the ABS.
This hypothesis was tested by correlating scores on the
efficacy scale with scores on the six Levels of the ABS.
Table 33 presents these data. As can be seen, there were
no significant results for Groups 1 and 3. For Group 2,
Levels 4, 5, and 6 (hypothetical, feelings, and reported
behavior) were significant at the .05 level. Combination
of Groups 1, 2, and 3 yielded significance at Levels 4 and
5 (hypothetical and feelings) and negative results for
Level 1 (stereotypic) at the .05 level of significance.

In Chapter VIII differences between Groups 1 and 2 are
further discussed, pointing out the 'efficacy-like' behavior
of Group 2. Why Group 1 seems different is not fully
understood. Some of these significant results may have
occurred by chance along (Type I error).

H-4: Subjects taking the Semantic Differential will
give the concept 'Negro person' a significantly
higher rating on the 'evaluation' dimension
than the concept 'black person.‘

Hypothesis 4 was tested by using a two-sample t-test
for dependent samples. Table 34 presents the results and
the t values required for the .05 level of significance. The
concept 'Negro person' was compared with the concept 'black
person' on the 'evaluation' dimension of the Semantic Diff—

erential. The results do not support hypothesis 4.

232

TABLE 33.--Correlations and Significance Levelsa of Efficacy
to the Six Levels of ABS: WB-WN

 

 

Level Group 1 Group 2 Group 3 Totals
N=22 N=25 N=253 N=300

l -21 (32) -13 (49) -ll (09) -ll (05)

2 09 (67) —ll (57) -Ol (82) 007 (89)

3 08 (68) 19 (34) 07 (29) 08 (20)

4 15 (49) 51 (006) 08 (19) 12 (04)

5 21 (31) 39 (04) ll (08) 15 (01)

6 -29 (16) 41 (03) ll (07) 09 (ll)

 

a . . . .
Significance Levels in parentheses.

TABLE 34.--t-Score Values for Semantic Differential Ratings
of 'Negro Person' and 'Black Person'

 

 

 

 

Evaluation Potency Activity
Group 1, N = 22
0.39 -0.43 -O.76
Group 2, N = 25
-0.04 -O.86 0.57

 

Required t—values for the .05 level of significance.
Group 1: N=22

t.05;21df = 1.72
Group 2: N=25

t.05;24df = 1.71

H-S: Subjects taking the Semantic Differential will
give the concept 'black person' a significantly
higher rating on the 'potency' dimension than
the concept 'Negro person.‘

This hypothesis was evaluated with a two-sample

t-test for dependent samples. The results are contained in
Table 34 and the t values required for the .05 level of

significance. The hypothesis was not confirmed; the nega-

tive results are also in direct contradition to the stated

233

hypothesis. The results, however, do not reach statistical
significance.
TABLE 35.--Correlations and Significance Levelsa of Efficacy

and Five Concepts of the Semantic Differential on
the Evaluation Dimension

 

Friend Enemy Black White Negro

 

Group 1, N=22

 

-27 (19) 02 (91) ~01 (94) 08 (71) 23 (27)

Group 2, N=25

 

68 (0001) -41 (03) 23 (24) 19 (33) 16 (33)

 

a . . . .
Significance levels in parentheses.

H-6: Subjects taking the Semantic Differential will
give the concept 'black person' a significantly
higher rating on the 'activity' dimension than
the concept 'Negro person.‘

Hypothesis 6 was tested using a two-sample t-test

for dependent samples. Table 34 presents the results. For
the two groups, the hypothesis was not confirmed; for Group
1, negative, but not significant results were found.

H-7: A higher efficacy score on the ABS will be
correlated with more positive scores on the
'evaluation' dimension of the Semantic Differ-
ential for the concepts 'friend,' 'black
person,‘ 'white person,‘ 'Negro person,‘ and
more negative scores for the concept 'enemy.‘

For Group 1, the hypothesis was not confirmed for

any of the five concepts evaluated. For Group 2, the hypothe-

sis was confirmed (Table 35) for two of the five concepts

in the direction predicted. The concept 'friend' correlated

234

positively (at the .05 significance level) with efficacy, and
the concept 'enemy' correlated significantly in the negative
direction with efficacy. The other three concepts, 'black

person,‘ 'white person,‘ and 'Negro person, did not corre-

late with the efficacy variable.

H—8: There will be a positive relationship between
high scores on the 'evaluation' dimension of
the concept 'black person' and 'Negro person'
of the Semantic Differential and positive
scores on the ABS.

As can be seen in Table 36 the hypothesis was not
confirmed. There were some significant results between the
two Semantic Differential concepts and some of the attitude
Levels. The two groups, however, offset each other. Where
significance was found between a Level and a concept in
one group, it was not found in the other group. If the two
groups' results were combined, significance would not be
evident. Also, the results may be due to Type I error,
finding significance where it may not actually be present.
It can be concluded that no relationship was found in this
study between the Semantic Differential concepts on the
'evaluation'dimension and the six Levels of the ABS.

H-9: There will be no significant correlations
between high scores on the 'potency' dimension
of the concepts 'black person' and 'Negro
person' of the Semantic Differential and
positive attitude scores on the ABS.

Hypothesis 9 was confirmed as shown in Table 37.

Some positive significance was found between particular

235

TABLE 36.——Corre1ation Coefficients and Significance Levelsa
Between the Six Levels of the ABS and the 'Evalu-
ation' Dimension of 'Black' and 'Negro' of the
Semantic Differential.

 

 

 

 

Group 1 N=22 Group 2 N=25
Levels

Black Negro Black Negro
l. Stereotypic 04 (84) 01 (95) -3O (12) -32 (10)
2. Normative -04 (85) -17 (42) 14 (47) 12 (56)
3. Moral Evaluation 45 (02) 28 (18) 01 (99) 10 (64)
4. Hypothetical Action 53 (008) 44 (03) 20 (32) 21 (29)
5. Feeling 18 (40) 20 (35) 31 (ll) 32 (10)
6. Action 13 (54) ll (61) 37 (05) 37 (05)

 

a . . . . .
Significance levels in parentheSis.

TABLE 37.—-Correlation Coefficients andSignificanceLeevelsa
Between the Six Levels of the ABS and the 'Potency'
Dimension of 'Black' and 'Negro' of the Semantic
Differential.

 

 

 

 

Group 1 N=22 Group 2 N=25
Levels

Black Negro Black Negro
l. Stereotypic 06 (78) 04 (84) -22 (27) -32 (10)
2. Normative ~34 (10) -25 (23) -25 (22) 30 (12)
3. Moral Evaluation -30 (14) -05 (82) 13 (51) O6 (75)
4. Hypothetical Action 38 (06) 39 (05) 25 (21) -02 (90)
5. Feeling 12 (56) 13 (54) 41 (03) 09 (64)
6. Action 21 (32) 38 (06) 44 (O2) -06 (76)

 

fl.

a I I O O I
Significance levels in parentheSis.

TABLE 38.--Corre1ation Coefficients and Significance Levelsa
Between the Six Levels of the ABS and the 'Activity'
Dimension of 'Black' and 'Negro' of the Semantic
Differential.

 

 

 

 

Group 1 N=22 Group 2 N=25
Levels

Black Negro Black Negro
l. Stereotypic ~11 (62) 19 (37) -33 (09) -38 (05)
2. Normative 02 (92) 14 (50) 29 (13) 28 (15)
3. Moral Evaluation 35 (09) 16 (46) -O6 (75) 13 (50)
4. Hypothetical Action 39 (06) 36 (08) 16 (41) 34 (08)
5. Feeling 32 (13) 22 (30) 15 (47) 20 (30)
6. Action 08 (70) 13 (54) 18 (36) 35 (07)

 

a . . . . .
Significance levels in parentheSis.

236

concepts and Levels, but this was only intermittent. Also,
it was not consistently found between the two groups tested.
H-lO: There will be no significant correlations
between high scores on the 'activity' dimen-
sion of the concepts 'black person' and
'Negro person' of the Semantic Differential
and positive attitudes on the ABS.
Hypothesis 10 was confirmed; no significance was
found (Table 38) between the six Levels of the ABS and the
concepts of 'black person' and 'Negro person' of the

Semantic Differential. Some significance was found between

certain terms and Levels.

CHAPTER VIII

SUMMARY, DISCUSSION, AND RECOMMENDATIONS

Democracy will not come
Today, this year

Nor ever
Through compromise and fear.

I have as much right

As the other fellow has
To stand

On my two feet

And own the land.

I tire so of hearing peOple say,

LET THINGS TAKE THEIR COURSE.

TOMORROW IS ANOTHER DAY.

I do not need my freedom when I'm dead.
I cannot live on tomorrow's bread.

Freedom
Is a strong seed
Planted
In a great need.
I live here, too.
I want freedom
Just as you
(Lanston Hughes, 1966, p. 285).

This chapter will briefly review the purpose of the
study; will summarize the main points stressed in the review
of the literature; and will state the results of the data
and hypotheses. Lastly, the implications and recommenda-

tions for futher research will be discussed.

237

238

Summary of the Study

 

Purpose

One of the major aims of this study was to discover
whether or not the two terms, 'black' and 'Negro,' elicit
different "attitude behaviors." Two scales, the Semantic

Differential and the racial Attitude Behavior Scale, were

 

used for evaluation of this hypothesis. A second purpose
was to review the theoretical basis of the ABS, exploring
psychological measurement and specific areas of attitude
research that led to the development of this scale. The
review of attitude research contained within this thesis

serves as introductory material for discussion of the ABS.

Literature

 

A review of attitude research prior to Guttman-
Jordan "attitude-behavior" theory and facet analysis was
conducted. Beginning with early Greek philosophy, the
statistical basis for attitude research was traced through
Greek Cosmologic and sophist theory, British Empiricism
and German Experimentalism, the factor analytic techniques
of Spearman and Thurstone, and the facet analysis of
Guttman. The theoretical development of attitude research

'act'

began with Plato, and proceeded through the school of
psychology, Sully's classification of mental states, Ward's
work on the subject-object relationship, Allport's attitude

research, McGuire's tripartite attitude classification

239

scheme, and Jordan's research on "attitude-behaviors"

including his development of the series of Attitude Behavior

 

Scales (ABS). With the ABS approach, it has been possible

to analyze the facets of attitude and provide a more rigorous,
methodological framework for attitude research. The litera-
ture on race and race prejudice was reviewed, concentrating
on the areas of race labelling, racial attitudes, and the
language of prejudice. Lastly, a review of research on

race, color labelling, and the Semantic Differential was

undertaken.

Instrumentation and Methodology_

 

Jordan and Hamersma (1969) constructed a series of
attitude scales based on the facet methods of Guttman (1959).
The scale used in this study, ABS-WB/WN-G is one in this

series of scales. These Attitude Behavior Scales have been

 

applied to numerous 'minority' groups cross-culturally

(Bray, 1972; Jordan, 1972; Smith, 1973).

Design and Analysis

The ABS-WB/WN-G scales were administered to three
groups of college students at Michigan State University.
These students were college seniors majoring in Education,
planning a career in teaching. Two ABS scales were admin-
istered, the "ABS toward blacks" and the "ABS toward
Negroes." One large group of 254 students was tested

Winter Term of 1972; two other smaller groups (samples of

240

22 and 25) were tested Spring Term of 1972. These two

latter groups were also given a Semantic Differential scale.
An analysis was performed comparing the results of the
Semantic Differential with the results of the ABS. Several
hypotheses were tested using correlation coefficients, mul-
tivariate analysis of variance procedures, two-sample t-tests,

and the Kaiser Q2 simplex approximation.

Research Findings

 

The results of the study indicated there were no
differences between the two groups, those taking the "ABS
toward blacks" and those taking the "ABS toward Negroes,"
on the six Levels of the ABS. The two concepts, 'black'
and 'Negro,' elicited the same "attitude-behaviors" as
measured by the ABS.

The attitude data from the three samples did
approximate a Guttman simplex. All groups and the com-
bined totals excelled Hamersma's (1969) .70 criterion g2
value required for approximating a simplex.

For the efficacy variable, significant relationships
were obtained in one group on Levels 4, 5, and 6 (Hypothet-
ical, Feeling, and Self-Reported Behavior); the other two
groups did not show significant relationships. For the
combined totals, some significant results were found
(Levels 4 and 5--Hypothetical and Feeling).

The hypotheses regarding differences between the

concept 'black person' and 'Negro person' on the 'evaluation,'

241

'potency,' and 'activity,' dimensions of the Semantic Differ-
ential were not confirmed.

Some correlations (two out of the five concepts of
the Semantic Differential) were found between efficacy on
the ABS and the 'evaluation' dimension of the Semantic
Differential for Group 2; for Group 1, no correlations were
found for any of the five concepts.

No significant correlations were found between the
'evaluation,' 'potency,' and 'activity' dimensions of the

Semantic Differential and the six Levels of the ABS.

Discussion

 

The following is a discussion of the results of the
hypotheses of the study.

H-l: Multivariate analysis of variance was used to
test this hypothesis for the three groups evaluated. The
results indicated that no differences were found between
the two terms, 'black' and 'Negro.' White students taking
the "ABS toward blacks" and those taking the "ABS toward
Negroes" responded in a similar manner on each of the six
Levels of the ABS. Thus, the concepts, 'black' and 'Negro,'
did not elicit different "attitude-behaviors" as measured
by the ABS. This finding is in contradiction to Williams'
(1966) study that found white students evaluating 'black
person' more negatively than 'Negro person' on a Semantic

Differential scale. The results of this present study using

242

the ABS are in accord with a more recent study conducted by
Lessing and Zagorin (1972), that found essentially no diff-
erences in evaluation of the two terms.

These contradictory findings may be due to a time
lapse between the studies conducted. Earlier studies that
did find a difference, were conducted at a time when 'Negro'
was the more accepted term applied to black people. The
black power movement was just beginning to surface at that
time, and much resentment and misunderstanding was expressed
by whites. At the present time, whites, especially the
student population, have become more aware of the Movement,
and perhaps are now beginning to accept the new terminology
of the black power movement. Whether acceptance of the
terminology of the black power movement means that whites
will or are beginning to accept more of "black culture"
itself cannot be concluded by this study. However, an
interesting speculation would be that acceptance of the term
'black' (and 'Afro-American') by whites may be an indication
of acceptance of "black culture."

The data from the present study [Lessing and Zagorin,
1972, p. 69], when viewed in the light of these
other studies, suggests that as of 1968-1969 the major
effect of black power ideology upon young people
was the removal of the odious connotations of black
without giving it clear and unequivocal priority
over other non-white ethnic designations.

H-2: The Guttman simplex approximation hypothesis

was confirmed. This conclusion was reached by both visual

inspection of the simplexes in Table 32 and utilization of

243

Kaiser's g2. Hamersma (1969) arrived at a Q2 value of .70
as the minimal level to conclude that data had approximated
a simplex. The possible range of scores is from 0.00 to
1.00. All the matrices in this present study exceeded the
.70 Q2 value. For the original matrices, the g2 values
ranged from .85 to .95. This is an improvement over the
other racial ABS simplexes. Dell Orto (1970) did not
obtain a simplex in his study of rehabilitation counselor
trainees; Frechette (1970) found simplexes ranging in

value from .54 to .91 in his study of French and English
speaking Canadians; Williams (1970) study of racial attitudes
of black and white policemen revealed simplexes from .73 to
.90.

In the present study, the directions in the instruc-
tions for some of the Levels had been modified; these
changes may explain why higher simplex scores were obtained.
These changes mainly concern the subject-object relation-
ship; the attempt was to make the Levels more understandable
to the individual taking the scale. Changes were made in
Levels 1, 2, 3, and 5. Level 6 was modified in the number
of item choices. Dell Orto used a four-response choice
option, the fourth choice being, "no experience." The
present scale omitted this "no eXperience" option in the
item choices. Instead, the directions specified that

subjects having no experience should omit the entire Level.

244

The changes made in Levels 1, 2, 3, and 5 are illustrated

below,

Level 1

i.e., from the WN-G version (Appendix B-2):

DELL ORTO SCALE:
This section contains statements about ideas which

most other whites have about Negroes. Circle or

 

 

fill in the answer sheet number that indicates how
other whites compare themselves with Negroes.

Please answer all questions.

Other Whites believe the following things about

 

Whites as compared to Negroes:
l. Whites can be trusted with money
1. more than Negroes
2. about the same as Negroes

3. less than Negroes

PRESENT SCALE:
This section contains statements about ideas which

other Whites have about Negroes. Circle or fill in

 

 

the answer sheet number that indicates how other
Whites compare Negroes with themselves. Please

answer all questions.

Other Whites believe the following things about

 

Negroes as compared to Whites:

Level 2

Level 3

245

l. Negroes can be trusted with money
1. less than Whites
2. about the same as Whites

3. more than Whites

DELL ORTO SCALE:
This section contains statements about things which

other Whites like you may believe about Negroes.

 

Please choose the answer that indicates what you

 

think others believe about Negroes.

 

Most Whites generally believe the following about

 

Negroes:

PRESENT SCALE:

This section contains statements about things which
most Whites generally believe others would experience
when interacting with Negroes. Please choose the

answer that_indicates what you think most others

 

believe about Negroes.

DELL ORTO SCALE:
This section contains statements about ways in which

ypupyourself should act toward Negroes. Please

 

choose the answer that indicates how you feel you

 

should act or believe.

 

Level 5

246

In respect to Negroes, do you yourself believe that

 

it is usually right or usually wrong:

PRESENT SCALE:
This section contains statements of the right or
wrong way of behaving or acting toward Negroes. You

are asked to indicate what you yourself believe

 

others think should be done with respect to Negroes.

In respect to Negroes, what do you believe others

 

think is right or wrong:

DELL ORTO SCALE:

This section contains actual feelings that White
people may have about Negroes. You are asked to
indicate how you feel about the following state-

ments.

How do ygu actuallypfeel toward Negroes:

 

PRESENT SCALE:

This section concerns actual feelings that you

 

yourself may have about Negroes. You are asked

to indicate how you feel about the following state-

 

ments.

How do you actually feel toward Negroes:

 

247

The large group of 254 students resulted in a 22 value of
.95; results for a combination of the three groups yielded

a Q of .90. According to Guttman's (1959) Contiguity
Hypothesis, this means that Levels closer to each other

were more related than Levels more distant. This simplex
approximation means that the semantic six-Level structure of
the ABS has structural relevance.

Thus, the ABS is one of the only attitude scales
currently in the literature that has formulated and instru-
mented a facetized definition of the concept, attitude. The
five-facet, six-Level design allows a researcher knowledge
of whether he is dealing primarily with the cognitive,
affective, or overt action (behavioral) component of atti-
tude. Other researchers (i.e., Allport, 1954; McGuire,
1969) have defined attitude as being composed of cognitive,
affective, and conative elements, but their definitions
were purely semantic and not instrumented. Jordan's (l971a)
system divides this tripartite classification into six
distinct Levels; the Contiguity Hypothesis and facet system
design gives the Levels structural and semantic relevance
to one another and an increasing proportion of strength
(cognitive to affective to conative/behavioral) as one
proceeds down the six Levels.

H-3: The efficacy variable purports to measure -
the amount of control a person feels he has over his environ-

ment. It was hypothesized that those who scored high on

248

the variable of efficacy would have more positive scores on
the six Levels of the ABS. No significant results were
found for Groups 1 and 3. However, significant findings
were evident for Group 2 on Levels 4, 5, and 6 (Hypothet-
ical, Feeling, and Self-reported Behavior). Also, when
the three groups were added together, significant results
were obtained for Levels 4 and 5. For Level 1, all groups
and the combined scores are in the Opposite (negative)
direction. The total of all these groups yields a score
close to the .05 level of significance, in the negative
direction. The positive scores on Levels 4 through 6 for
Group 2 and Levels 4 and 5 for the combined group totals
involve affective and conative material.

The negative results for Level 1 represent responses
to cognitive material. This Level is a measure of an indi-
vidual's responses to Stereotypic items on the Attitude

Behavior Scale. The lack of significant relationships at

 

Levels 2 and 3 and the negative scores for Level 1 may
indicate that the variable, efficacy, correlates with the
ABS only at the affective and conative Levels; that "sense
of control over one's environment" is related to feeling and
behavioral dimensions. Clinical observations supported

this notion for Group 2. This group was a much more inter-
ested, active, and responsive group, as Opposed to Group 1.
On appearance, there was definitely a marked difference

between these two groups. This interesting finding is

249

discussed more extensively under H-7 of this chapter. For
more liberal students (presumably Group 2), efficacy corre-
lates with the more affective and behavioral Levels of the
ABS, whereas for more conservative students (presumably

Group 1), efficacy does not correlate with the ABS Levels.
From the differences in the views expressed and the topics
chosen, it appeared that Group 2 was composed of more liberal
students than Group 1. It should also be noted that the
three black students in Group 2 helped stimulate discussion.
However, the two most vocal of the black students were not
present the first two days of class, and discussion was still
much more intense than in Group 1.

Frechette (1970) found a correlation between effi-
cacy and Levels 1 and 4 of the ABS. It can be conjectured
that his Canadian samples were more conservative than
Group 2 and more similar to the students in Group 1. To
test for this possibility, one might administer the ABS to
a sample of students known to be more liberal than these
groups tested.

Frechette (1970) found different results; his
administration of the racial ABS revealed that Levels 1 and
4 of the ABS correlated positively with efficacy, whereas
Levels 5 and 6 did not. He concluded that efficacy may
correlate with less behavioral and affective areas, and more

with cognitive material.

250

The present author interprets the ABS Levels in a
different manner than Frechette. Level 1, Stereotypic, is
indeed a measure of cognitive material; however, Level 4,
Personal Hypothetical Interaction, is primarily a measure in
tne affective area. With this interpretation, a portion of
Frechette's results are consistent with the present study.
His results for Level 4 parallel the results found on
positive correlation between Level 4 and the variable,
efficacy. The present author considers this correlation with
efficacy as one involving an affective Level on the ABS, not
a cognitive one, as Frechette believed.

Frechette also noted a positive correlation between
Level 1 and efficacy; the present author agrees that this
Level is a measure of cognitive responding. The present
study of college students found no correlation between the
efficacy variable and Level 1.

Several reasons may suggest why this difference was
found between the two studies. The pOpulations were
different. Frechette's sample was not only drawn from a
different country (Canada), but involved distinct urban
areas (Toronto and Montreal). The sample was considerably
older and had several years working experience in the
occupational world. The present study used a sample of
college seniors majoring in education in a large midwestern

university, a sample that can be considered less urban than

Frechette's. Also, the attitude object was different in

the two studies; Frechette's study investigated attitudes of
French and English speaking Canadians toward West Indian
Immigrants, while the current study investigated white
attitudes toward blacks in the United States.

Lastly, these differing results indicated that those
in the Canadian sample tested who have a high degree of
efficacy have a clearer conception of what prejudices
their society has toward West Indian Immigrants than do
those with a lower degree of efficacy. No differences were
found for those who scored high or low on efficacy and
Level 1 of the ABS in the United States college sample,
assessing white attitudes towards blacks.

Since the data are correlations, inferences as to
cause and effect cannot be made. One can only state that,
in the present study, a positive relationship was found
between efficacy and Levels, 4, 5, and 6 on the racial ABS.

H-4: White students did not rate 'black person'
more negatively than 'Negro person' on the 'evaluation'
factor of the Semantic Differential. This supports Lessing
and Zagorin's (1972) argument that the words 'black' and
'Negro' no longer represent different connotations to white
college students. Both their results and the results of
the present study indicated that white college students no
longer preferred the term 'Negro' over 'black' when referring

to black people. Jordan and Hamersma (1969) found that

252

whites preferred 'Negro' over 'black' in the late 1960's.
Williams' (1966) study indicated that white college students
had rated 'Negro' higher in acceptance than 'black.‘ The
time factor between the two sets of studies may have played
a large part in the differences found. Whereas, in the
1960's the term 'black' was still relatively unacceptable

by whites; by the early 1970's, whites may have become more
familiar with the term, and hence, more accepting of it.

H-5 and H-6: The results from Hypotheses 5 and 6,
indicate that white college students do not differentiate
between the concepts 'black person' and 'Negro person' on
the 'potency' and 'activity' dimensions of the Semantic
Differential. Lessing and Zagorin (1972) stated the following
results for both black and white college students: "Blacks
were seen as the strongest, most potent, and most active.
Negroes were viewed as slightly less strong and active . . ."
(p. 70). The results of the present study found just the
opposite for 'potency' and 'activity.’ 'Negro person' was
evaluated slightly higher on the 'potency' and 'activity'
dimensions of the Semantic Differential than 'black person.‘
The results showed this trend in direction, but were not
found to be significant.

The black power movement has now been present for
several years. Perhaps this exposure, especially prevalent
on the college campuses, has become more accepted by white

students; they no longer differentiate between the concepts,

'black' and 'Negro.'

The differences in the samples tested may also have
had an effect. Lessing and Zagorin (1972) tested under-
graduate students in five midwestern colleges and universi-
ties. The present study tested seniors majoring in educa-
tion in a large midwestern university.

There was a time difference between this present
study and the one conducted by Lessing and Zagorin (1972).
They collected their data in 1968-1969, while the data
for the present study was gathered in 1972. Many other
factors, besides the black power movement, have occurred
in this time period that could influence the results.

The testing in the Lessing and Zagorin study was
conducted by a biracial team or by a black professor. The
present study used white examiners. This difference may
have played a part in influencing the results of the two
studies.

H-7: Hypothesis 7 was partially confirmed for
Group 2 and not confirmed for Group 1. Group 2's evaluation
of 'friend' correlated positively with the efficacy variable,
and 'enemy' correlated negatively with efficacy. Both these
results were as predicted. This group's evaluation of the
three other concepts, 'black,‘ 'white,' and 'Negro,' did
not correlate with efficacy.

Clinically, there was a noticeable difference
between Groups 1 and 2. Both groups were sections of

Education 450, titled, "Attitudes Toward Minority Groups."

254

The first section was a very quiet, seemingly uninterested
group, while the second section had many spontaneous and
sometimes quite heated discussions concerning minorities,
prejudice, and racism. It is felt by this observer, that
Group 2 was more interested in studying, learning about,
and discussing minority problems than was Group 1. It is
possible that there were some selectivity variables that
made the two sections different. Section 2, the more
active group, was held at 12:20 in the afternoon, while
Section 1 met at 8:00 a.m. Perhaps the students that were
interested in the subject matter signed up early, choosing
the afternoon section over an early morning section. More
students in the morning section indicated that it was one
of the only sections left (out of 74 'modules' offered,
covering various topics) when they signed up.

Efficacy is a measure of a person's feelings of
control over his surrounding environment. Perhaps those
students who felt more in control of their surroundings
(high in efficacy) came earlier to sign up for the section
of their choice, and were able to choose the afternoon
section. Those students lower in efficacy may have signed
up late when only a few 'modules,' including the 8:00 a.m.
"Attitudes Toward Minority Groups" section, were not filled.
Since this analysis is based on clinical observations, addi-
tional studies controlling for intentionality of choice

(voluntary versus no choice) need to be administered.

255

The significance of these results must remain ten-
tative. Chance significance (Type I error) may have pro-
duced a greater correlation than actually occurred.

H-8: Apparently, the ABS and the 'evaluation'
dimension of the Semantic Differential are not measuring
the same thing. Both scales claim to be measuring attitude.
If this is true, then the two scales are measuring different
aspects of attitude. The Semantic Differential scale is
concerned with semantic meanings of words: "score variation
along the E 'evaluation' dimension covaries closely with the
score variation on conventional attitude tests" according
to Osgood, Suci, and Tannenbaum (1957, p. 193). Jordan's

Attitude Behavior Scale is a relatively new attitude

 

scale. It is based less on conventional attitude tests
and judgments of 'experts' in the areas of race and race
relations and more on Guttman's facet analysis.

According to Fishbein (1965), Osgood defined
attitude as a "prediSposition to respond;" Jordan (l971a)
on the other hand, defined "attitude as behavior," and
stated that attitude was not a "predisposition to respond."
Thus, the authors of the two scales defined attitude in
two different manners. It is difficult to identify exactly
what the Semantic Differential is measuring. Jordan's
system of "attitude-behavior" is based on a five-facet,
six-Level system; a structural system underlies his seman-

tic definition of attitude. It is conceivable that the two

256

scales, the Attitude Behavior Scale and the Semantic Differ-

 

ential, are simply not comparable instruments. The impli-
cations of this are discussed in the "Recommendations for
Further Research" section of this chapter.

H-9 and H-lO: It was speculated that the 'evalua-
tion' factor of the Semantic Differential would correlate
positively with the six Levels of the ABS. Further, it was
speculated that if these two scales correlated on a measure
of attitude ('evaluation' on the Semantic Differential and
the six Levels of the ABS), the two additional measures of
the Semantic Differential, 'potency' and 'activity,' would
not necessarily correlate with the ABS Levels. It was
hypothesized that 'potency' and 'activity' would not corre-
late with the ABS because the authors of the Semantic
Differential felt that 'evaluation' was a strong measure
of attitudes, but 'potency' and 'activity' were much
weaker measures of attitude. Hypotheses 9 and 10 were
confirmed; no correlation was found between 'potency' and
'activity' on the Semantic Differential and the six Levels

of the ABS.
Recommendations for Further Research

'Black' Versus 'Negro'
The present data are limited to white senior edu-
cation majors in a particular, large midwestern university.

The study by Lessing and Zagorin (1972) was also conducted

257

in midwestern universities. It is believed that although
these results may generalize to other American universities
and colleges, different results might be found in non-
college populations. If older persons (particularly those
without a college education) were given the scales, they
might respond more positively to the term 'Negro' than to
the term 'black.‘ This speculation is raised because before
black power ideology was a prominant issue, college students
responded more positively to the word 'Negro' than to 'black'
(Williams, 1966). One possibility is that the black power
movement, stressing the slogan "black is beautiful," has
had an effect in decreasing the negative connotations of
the word 'black.‘ A second possibility is that white
college students are beginning to accept the black power
and culture movement, and the notion that black people prefer
to be called 'black' rather than 'Negro.' In 1969, Jordan
and Hamersma felt that the term 'black' was more acceptable
to the black population, but that 'Negro' was still preferred
by whites. It appears that the word 'black' can now be
used when testing white college populations.

It would be interesting to give the two ABS scales
to a black population to test how they react to the word
'black' versus 'Negro.' This author would conjecture that
the word 'black' would elicit more positive attitudes on

the ABS than would the word 'Negro' among black people.

258

This result has a broader implication when viewed
with the implications concerning color labelling and word
connotations. Throughout history, the word black has always
had negative connotations associated with it; recent studies
point out that these negative associations still exist,
contrasted with positive associations for the word white.
The words, black and white, are often seen as polar opposites.

The black power movement, in this country, has been
promoting the theme, "black is beautiful." The trend has
been toward elimination of the word Negro in reference to
black peOple. In place of Negro, black power advocates have
suggested use of such alternatives as: Afro-American,
Afram, African American, and black, Williams (1970) suggested
that since the term black has been so filled with negative
connotations, perhaps it would be beneficial to eliminate
reference to racial groups by color names.

Results of the present study indicated that white
college students do not necessarily associate the negative
references of the word black with persons so designated as
black. The concept "black person" did not elicit any more
negative evaluations than "Negro person" on the two attitude
instruments. Therefore, it seems that use of the term
black in reference to the black race, will not necessarily
carry with it the negative connotations associated with the

word black.

259

Efficacy and the ABS

 

Clinical observation of the two Spring Term sections
of "Attitudes Toward Minority Groups" raised some very
interesting questions. Group 2 was much more involved with
discussion of minority problems and prejudice. Their
interest level was high throughout each individual classroom
session and throughout the three weeks of class. Group 1,
on the other hand, appeared more to be just "putting their
time in" and little interest or discussion could be generated
in this section. At the same time, there was a significant
difference between correlations of efficacy scores and the
ABS. Group 2's scores on efficacy correlated positively
with the three "feeling and action" Levels of the ABS
(Levels 4, 5, and 6) while Group 1's scores did not corre—
late with any of the six ABS Levels. The correlations
obtained in GrOUp 2, may indicate that those who feel in
control of their environment also score positively on
feeling and action levels of racial "attitude-behaviors."

It can be postulated that individuals in Group 2
were there more because they voluntarily chose to be, while
those in Group 1 were there because of a "no other choice
position." The presence of three black students in Section
2 also helps confirm this. Lastly, more students in Group

"no

1, when asked why they chose this section, stated
choice," than students in Group 2. Since this is only a

clinical speculation and voluntariness of choice was not

260

controlled for, a future study with the racial ABS might

seek to control for voluntariness and/or interest level and
find out if those individuals more interested in racial
issues, score higher on efficacy and have more positive
"attitude-behaviors" on the affective and conative Levels

of the ABS. This variable needs additional study, especially
since an earlier study with the racial ABS (Frechette, 1970)
revealed opposite results. Frechette found that the variable,
efficacy, was significantly related to Levels 1 and 4, on

the ABS.

The Simplex Approximation and the ABS

The simplex approximation for the racial ABS in
this study, was high enough to state a semantic and structural
relationship between the six Levels. Jordan and his asso-
ciates have postulated a definition of "attitude-behavior"
that is founded on a semantic structure. It was first
necessary to arrive at a definition for "attitude-behavior;"
the next step is to include the concept of "attitude-behavior
change" into this approach. The definition of "attitude-
behavior" can be made more relevant if the notion of change
is included. Basically, what is needed is a monor redis-
tribution-modification of Jordan's five-facet, six-Level
theory.

The following is a proposal to include "attitude-
behavior change inclination" into the present facet analysis
of "attitude-behavior." The five-facet, six-Level system

should be maintained within this present modification.

261

Levels 1 through 3 might be combined into two Levels, or

one of these Levels might be eliminated. Levels 4 through

6 are the important Levels, since they deal with affective and
conative behavior, and should be retained as they are
presently defined. With the elimination of one of the first
three Levels or modification thereof, an 'additional' Level
can be added without destruction of the basic structural
design.

This additional Level would attempt to measure
"attitude-behavior change inclination"--self-reported areas
of change concerning racial interaction that would be 'most'
acceptable and 'least' acceptable to change. In other words
(for the racial scale),what areas would be acceptable for
a particular individual to change that would bring him into
closer contact and/or relationships with the 'race' in
question. The directions and items for this Level would
encompass the following:

I would like the opportunity to work with blacks;
to move into a racially integrated neighborhood;
etc.

This additional Level would contain two sets of
questions, the second set to include and differentiate those
who have had the eXperiences. This second question (imme-
diately following each item) would tap satisfaction with the
experience. The individual with experience would be asked

whether he was (a) dissatisfied, (b) uncertain, or

262

(c) satisfied, with the particular experience. This pro-
cedure would aid in detecting those who have already had
the eXperience, what their perceptions of the experience
were, and what percent of the sample being tested have had
the experiences.

The facets would have to undergo some modification
for the remaining five Levels to encompass the newly modi-
fied Level within the five-facet design. Thus, the ABS
would be obtaining data in areas of resistance to and accep-
tance of "attitude-behavior change." Researchers would then
have a better idea as to what areas were reported as being
"Open to change" and the degree of resistance that might
be encountered for each issue. Areas of 'agreeable' change
could be readily implemented. Areas of high resistance
could be investigated further before attempting change.
Needed programs could be implemented in the difficult areas
for preliminary study before change was attempted. Another

possibility is that as the areas of least resistance are

 

implemented, i.e., integrated jobs, integrated colleges,

areas of more resistance, i.e., integrated high schools,

 

integrated housing, would become easier to implement.

An example of this lies in the area of the recent
busing controversy. If the attitudes of people toward busing
to achieve racial integration of schools could have been
evaluated (i.e., via the new Level), many of the complications

that arose may have been reduced or eliminated. If it had

263

been known that so much resistance would be encountered,
alternatives could have been presented. This is not to say
that measures such as busing should not be undertaken, only
that each action program should have a good chance in
succeeding to change "attitude-behaviors" in a positive

direction.

Additional Uses of the ABS

 

The ABS is too long an instrument; some restlessness
was evidenced in subjects taking the instrument. Elimina-
tion of items from each Level is recommended. Item analysis
could be run on the items, retaining the better items. This
might also improve the simplex.

The ABS should be used in experimental situations.
It could be used 'pre' and 'post' to measure amount of atti-
tude change produced in various classroom and/or eXperience
situations.

Longitudinal studies can be conducted with the ABS.
"Attitude-behaviors" could be tested yearly on certain
groups to assess the amount of change occurring. This could
be done in a high school or college, as students proceed
through an educational program.

The racial ABS can be used to test attitudes among
different populations. Studies could be conducted in
different geographical regions of the United States; popu-

lations such as rural groups, differing age groups, groups

264

from different income strata, and samples from different
ecnomic levels could be sampled. Also, the scale should be
administered to some all black groups of subjects to evaluate
racial attitude-behaviors of blacks toward whites.

The ABS and the Semantic
Differential Compared

 

 

The major differences between the ABS and the Seman-
tic Differential can be found both in the actual design of
the scales and in the two authors' definitions of attitude.

Osgood, Suci, and Tannenbaum (1957, p. 9) saw the
Semantic Differential as relating "to the functioning of
representational processes in language behavior" and as
serving as an index of these processes. These authors were
more concerned with the language and verbal aspects of
attitude than with behavioral dimensions. Also, their scale
does not involve specific situations, the attitude object
in-situation, but rather asks an individual to rate and
evaluate the perceived goodness or badness of a concept
as it relates to a group of people or idea.

The ABS, on the other hand, is more concerned with
the behavioral dynamics of attitude; the author of the scale
views attitude and behavior as one concept, hence, his use
of the term, "attitude-behavior." He defines attitude-
behavior along six Levels, from cognitive to affective to
conative aspects. Also, the ABS differs from the Semantic

Differential in that the ABS asks the subject to evaluate

265

specific situations that he may have encountered in inter-
acting with the particular group in question. This
involves evaluation of the attitude-object toward a situation,
while the Semantic Differential involves evaluation of an
attitude-object in a more global or undefined space.

This is not to say that either instrument fails to
measure attitude. Each author defined attitude as he saw
it and then set out to measure attitude as he had defined
it. Each scale must be viewed as measuring attitude within
a particular context. If the data from two attitude scales
are to be compared with each other, they must at least have
similar semantic content via which to assess the concept
of attitude.

If two attitude instruments are used for the purpose
of comparison, they should be selected so that the authors

define and instrument similarly the concept, attitude.

§EEE2£1
l. The two forms of the ABS, "ABS Toward Blacks" and

"ABS Toward Negroes," did not elicit different
"attitude behaviors" as measured by the six Levels
of the scale. Implications concerning this result
were discussed.

2. The racial ABS and the Semantic Differential used in
this study do not correlate with one another; hence,
the two instruments do not seem to be measuring

the same 'aspects' of attitude.

266

3. Some correlations were found between the variable
of efficacy on the ABS and Levels 4, 5, and 6 on

the Attitude Behavior Scale. These results are

 

only tentative and further research is suggested
before definite conclusions can be reached.

4. The racial ABS data resulted in a good approximation
of the simplex semantic structure between the six
attitude-behavior Levels.

5. Suggestions for incorporating "attitude-behavior

change inclination" into the Attitude Behavior

 

Scales were discussed.

6. The study using the racial ABS could be replicated
on different groups of subjects: all blacks,
different geographical locations (southern states),
rural populations, different age groups, income
groups, and economic levels.

Of the methods presently available for assessing
and defining attitudes, Jordan's is one of the few that
attempts to assess attitudes on both a semantic and a statis-
tical or structural basis. One can now begin to use the
ABS approach both to study the structure of attitudes, the
determinants or origins of attitudes, and to suggest pro-
cedures for changing attitudes.

Community 'operationalism' is necessary if the
'theory' is to become relevant to the world beyond the

academic arena. Unless theories are applied to the community,

267

their value must be questioned. It is not enough to simply
devise a theoretical model or system--the theory must be
put to some practical use. Jordan and his associates have
defined both semantically and structurally what they see

as "attitude-behavior." The next step, proposed in this
thesis, is to incorporate "attitude-behavior change incli-
nations" into this theoretical system.

One can then use the theory to study attitudes and
attitude change of individuals in society. Where it is
apparent that negative attitudes exist that are detrimental
to other people, programs to change these attitudes can be

instrumented. The Attitude Behavior Scale can be used to

 

test whether or not attitude change is occurring and on what
Levels. Jordan's (l971a) research indicated that change in
cognitive Levels (Levels 1, 2, and 3) does not necessarily
produce positive overt behavioral changes. An example of
this is that giving additional information about race and
race prejudice in a classroom teaching situation does change
an individual's cognitive perceptions of other persons
attitudes but will not necessarily change his own overt
behavior.

To change personal overt behavior, changes must
occur in an individual's attitudes on Levels 4 through 6,
the affective and conative dimensions of attitude-behavior.
Applied programs to change negative attitudes must be

directed toward these latter Levels of attitude-behavior.

268

This involves setting up situations where members of diff-
erent races are interacting with one another behaviorally,
and also 'enjoying' or 'benefiting' from the contact. With-
out this personal interaction and enjoyment of it, it appears

that negative attitudes will remain resistant to change.

REFERENCES

269

REFERENCES

Adorno, T. W., Frenkel-Brunswick, Levinson, D. J., &
Sanford, R. N. The authoritarian personality.
New York: Harper & Row, 1950.

 

Allport, G. W. Attitudes. In C. M. Murchison (Ed.),
Handbook of social psychology. Worchester, Mass-
achusetts: Clark University Press, 1935, Pp. 798-
844.

 

Allport, G. W. The historical background of modern social
psychology. In G. Lindzey (Ed.), The handbook of
social psychology. Massachusetts: Addison-
Wesley, 1954, Vol. I, Pp. 3-56.

 

 

Anastasi, A. Psypholpgical testing. London: Macmillan,
1968.

 

Angeles, P. A. The possible dream toward understanding
the black experience. New York: Friendship Press,
1971.

 

 

Banks, J. A. & Joyce, W. W. Teaching social studies to the
culturally different children. Reading, Massachu-
setts: Addison-Wesley, 1971.

 

 

Barnhart, C. L. (Ed.) The American college dictionary.
New York: Random House, 1959.

 

Bastide, R. & van den Berghe, P. Stereotypes, norms and
interracial behavior in Sao Paulo, Brazil. American
Sociological Review, 1957, 22, 689-694.

 

Baughman, E. E. Black Americans: A psychological analysis.
New York: Academic Press, 1971.

 

Bennett, L., Jr. What's in a name? Ebony, November, 1967,
23, 46-54.

Bem, D. J. Attitudes as self-descriptions: Another look at
the attitude-behavior link. In A. G. Greenwald,
T. C. Brack & T, M. Ostrom (Ed's.), Psyphological
foundations of attitudes. New York: Academic
Press, 1968.

 

 

Ben-Sira, Z. & Guttman, L. A facet-theoretical approach to
research on the use of media and on attitudes among

270

271

elite groups. United States Information Service,
Grant No. lA—637—472, Jerusalem: The Israel Insti-
tute of Applied Social Research, 1971.

Blauner, R. The question of black culture. In J. F. Szwed
(Ed.), Black America. New York: Basic Books, 1970.

 

Bogardus, E. S. Measuring social distance. Journal of Applied

 

Sociology, 1925, 9, 299-308.

 

Bogardus, E. S. Racial distance changes in the United
States during the past thirty years. Sociology and
Sociolggical Research, 1958, 43, 127-135.

 

 

Borgatta, E. F., Kottrell, L. 8., Jr. & Mann, J. H. The
spectrum of individual interaction characteristics:

An inter-dimension analysis. Psychological Reports,
1958, 4, 279-319.

Boring, E. G. A history of experimental psyphology. New
York: Appleton-Century-Crofts, 1950.

 

Bray, D. H. Ethnic values and attitudes. In D. H. Bray and
C. G. N. Hill (Ed's), Polynesian and Pakeha in New
Zealand: The sharing of cultures. VoIT’I, Auckland,
New Zealand: Heinemann Educational Books, Ltd.,
1972.

 

 

Bray, D. H. Attitudes and values of Polynesian and Pakeha
in New Zealand: Social and educational implications
of research findings. In K. W. Thompson & A. D.
Turlin (Ed's.), Contemporary New Zealand: Human
resources and social problems. Palmerston North,
New Zealand: Hicks, Smith, and Sons, 1973, in
press.

 

 

Brink, W. J. & Harris, L. The Negro revolution in America.
New York: Simon and Schuster, 1964.

 

Brink, W. J. & Harris, L. Black and White: A study of
U. S. racial attitudes today. New York: Simon and
Schuster, 1967.

 

 

Brinton, J. E. Deriving an attitude scale from semantic
differential data. In J. G. Snider & C. E. Osgood
(Ed's.), Semantic differential technique: A source-
book. Chicago: Aldine, 1969, Pp. 467-473.

 

Brown, T., Jr. Black is. New York: Grove Press, 1969.

Bryant, E. G., Gardner, 1., Jr. & Goldman, M. Responses on
racial attitudes as affected by interviewers of
different ethnic groups. Journal of Social Psycho-
logy, 1966, 70, 95-100. I

 

272

Burt, C. Experimental tests of general intelligence.
British Journal of Psychology, 1909, 3, 94-177.

Campbell, A. Civil rights and the vote for president.
Psychology Today, 1968, 1, 26-31.

 

Campbell, A. White attitudes toward blackppeople. Ann

 

 

 

Arbor, Michigan: Institute for Social Research,
1971.

Campbell, A. & Shuman, H. Racial attitudes in fifteen
American cities. Supplemental studies for the
National Advisory Commission on Civil Disorders.
Washington, D. C.: Government Printing Office,
June, 1968.

Cattell, R. B. Factor analysis: An introduction and manual

 

for the psychologist and social scientist. New
York: Harper and Brothers, 1952.

 

CBS News. White and Negro attitudes toward race related
issues and activities. New York: CBS, 1968.

 

Chaplin, J. P. & Krawiec, T. S. Systems and theories of
psychology. New York: Holt, Rinehart, and
Winston, 1968.

 

 

Citron, A. F. The 'Rightness of Whiteness': The world of
the White child in a segregated society. Detroit,
Michigan: Michigan-Ohio Regional Educational
Laboratory, 1969.

 

 

Clark, J. Manual of computerpprograms. East Lansing:
Research Services, Department of Communications
Michigan State University, 1964. '

Clark, K. B. & Clark, M. P. Racial identification and pre-
ference in Negro children. In T. M. Newcombe &
E. L. Hartley (Ed's.), Readings in social psycholpgy.
New York: Holt, 1947, Pp. 169-178.

 

Clarke, J. H. Black power and black history. Negro Digest,
1969, 18, 13-31.

 

Crombie, A. C. Apgustine to Galileo: The history of
science, A. D. 400-1650. London: Falcon Press,
Ltd., 1952.

 

 

Dell Orto, A. E. A Guttman facet analysis of the racial
attitudes of rehabilitation counselor trainees.
Unpublished doctoral dissertation, Michigan State
University, 1970.

273

Detroit Free Press. Do most people care about the next guy?
July, 21, 1972, p. l-C.

 

Dubois, W. E. B. The souls of black folk. New York:
Fawcett World Library, 1961.

 

Ebony Photo-editorial, What's in a name? Ebony, June, 1968,
23, 164.

Erb, D. L. Racial attitudes and empathy: A Guttman facet
theory examination of their relationships and deter—
minants. Unpublished doctoral dissertation, Michigan
State University, 1969.

Fanon, F. Black skins, white masks. New York: Grove Press,
1967.

 

Féré, C. Note sur la physiologie de l'attention. Review
of Philosophy, 1890, 30, 393-405.

 

Finn, J. D. Multivariance--univariate and multivariate
analysis of variance and covariance: A FORTRAN IV
ro ram. Occasional paper No. 9, School for Advanced
Studies, College of Education, Michigan State
University, 1970.

Fishbein, M. A consideration of beliefs, attitudes, and
their relationship. In I. D. Steiner and M. Fish—
bein (Ed's.), Current studies in social psychology.
New York: Holt, Rinehart and Winston, 1965, Pp.
107-120.

Fishbein, M. A. & Raven, B. H. The AB scales: An opera-
tional definition of belief and attitude. Human
Relations, 1962, 15, 35-44.

 

Foa, U. G. The contiguity principle in the structure of
interpersonal relations. Human Relations, 1958,
11, 229-238.

 

Foa, U. G. Three kinds of behavioral change. Psychological
Bulletin, 1968, 70, 460-473.

Foskett, D. J. Classification and indexing in the social
sciences. London: Butterworth Co., I963.

Frechette, E. J. Attitudes of French and English speaking
Canadians toward West Indian Immigrants: A Guttman
facet analysis. Unpublished doctoral dissertation,
Michigan State University, 1970.

Goldman, P. Report from Black America. New York: Simon
and Schuster, 1970.

274

Goldschmid, M. L. Black Americans and white racism. New
York: Holt, Rinehart and Winston, 1970.

 

Gottlieb, K. R. A Guttman facet analysis of attitudes
toward mental retardation in Colombia: Content,
structure and determinants. Unpublished doctoral
dissertation, Michigan State University, 1973.

Green, B. F. Attitude measurement. In G. Lindzey (Ed.),
The handbook of socialppsychology. Reading,
Massachusetts: Addison-Wesley, 1954, Vol. I,

Pp. 335-369.

 

Greenberg, H. M. Development of an integration attitude
scale. Journal of Social Psychology, 1961, 14,
103-109.

 

Guilford, J. P. Psychometric methods. New York: McGraw-
Hill, 1954.

 

Gunnings, T. S. Preparing the new counselor. The Counsel-
ing Psychologist, 1971, 2, 100-101.

 

 

Guttman, L. A basis for scaling qualitative data. Ameri-
can Sociological Review, 1944, 9, 139-150.

 

Guttman, L. The problem of attitude and opinion measure-
ment. In S. A. Stouffer, et al., (Ed's.), Measure-
ment and_prediction. New York: John Wiley and
Sons, 1950, Pp. 46-59 (a).

 

Guttman, L. The basis for scalogram analysis. In S. A.
Stouffer, et al., (Ed's.), Measurement and predic-
tion. New York: John Wiley and Sons, 1950, Pp.
60-90 (h).

 

 

Guttman, L. Image theory for the structure of quantitative
variates. Psyphometrika, 18, 1953, 277-296 (a).

 

Guttman, L. A note on Sir Cyril Burt's factorial analysis
of qualitative data. British Journal of Statisti-
cal PsycholOQYI 1953, 6, 1-4 (b).

 

 

Guttman, L. The principle components of scalable attitudes.
In P. F. Lazarsfeld (Ed.), Mathematical thinking
in the social sciences. Illifiois: Free Press, 1954,
Pp. 216-257 (a).

 

 

Guttman, L. A new approach to factor analysis: The radex.
In P. F. Lazarsfeld (Ed.), Mathematical thinking in
the social sciences. Illinois: Free Press, 1954,
Pp. 258-348 (b).

 

 

275

Guttman, L. An outline of some new methodology for social
research. Public Opinion Quarterly, 1954-55, 18,
395-404.

 

Guttman, L. What lies ahead for factor analysis? Educa-
tional and Psychological Measurement, 1958, 18,
497-515.

 

Guttman, L. A structural theory for intergroup beliefs
and action. American Sociological Review, 1959,
24, 318-328.

 

Guttman, L. Order analysis of correlation matrices. In
R. B. Cattell (Ed.), Handbook of multivariate
experimental psyohology. Chicago: Rand McNally,
1966, Pp, 438-458.

 

 

Guttman, L. Meeting of experts on race and race prejudice.
Paper presented at UNESCO, Paris, September, 1967,
mimeo.

Guttman, L. Measurement as structural theory. Psychome-
trika, 1971, 36, 329-347.

Guttman, L. & Schlesinger, I. M. Development of diagnostic
analytical and mechanical ability tests through
facet design and analysis. Cooperative Research
Program of the Office of Education, U. S. Depart-
ment of Health, Education and Welfare, Grant No.
OE-4-21-014, Jerusalem: The Israel Institute of
Applied Social Research, 1966.

Guttman, L. & Schlesinger, I. M. Systematic construction
of distractors for ability and achievement test
items. Educational and Psychological Measurement,
1967, 27, 569-580.

 

Hamersma, R. J. Construction of an attitude-behavior scale
of Negroes and whites toward each other using Gutt-
man analysis. Unpublished doctoral dissertation,
Michigan State University, 1969.

Hamersma, R. J., Paige, J. & Jordan, J. E. Construction of
a Guttman facet designed cross-cultural attitude-
behavior scale toward racial—ethnic interaction.
Educational and Psychological Measurement, 1974, in
press.

 

Hamilton, C. V. An advocate of black power defines it
(1968). In T. Wagstaff (Ed.), Black power: The
radical response to white America. Beverly HilIs,
California: Glencoe Press, 1969.

 

 

276

Harden, J. Response. Ebony, January, 1968, 23, 10.

Harding, J. & Hogrefe, R. Attitudes of white department
store employees toward Negro co-workers. Journal
of Social Issues, 1952, 8, 18-28.

 

Harrelson, L. E. A Guttman facet analysis of attitudes
toward the mentally retarded in the Federal Republic
of Germany: Content, structure and determinants.
Unpublished doctoral dissertation, Michigan State
University, 1970.

Harrelson, L. E., Jordan, J. E. & Horn, H. An application
of Guttman facet theory to the study of attitudes
toward the mentally retarded in Germany. Journal
of Psychology, 1972, 80, 323-335.

 

Heidbreder, E. F. Seven psychologies. New York: D.
Appleton Century, 1933.

 

Herring, J. P. Verbal and abstract elements in intelligence
examinations. Journal of Educational Psychology,
1921, 12, 511-517.

 

Hinckley, E. D. The influence of individual opinion on
construction of our attitude scale. Journal of
Social Psychology, 1932, 3, 283-296.

 

 

Hovland, C. I. & Sherif, M. Judgmental phenomena and scales
of attitude measurement: Item displacement in
Thurstone scales. Journal of Abnormal and Social
Psychology, 1952, 47, 822-823.

 

 

Hoyt, C. J. Test reliability estimated by analysis of
variance. In W. Meherns & R. Ebel (Ed's.), Prin-
ciples of educational and psychological measurement.
ChiCago: Rand McNally, 1967.

 

Hughes, L. Selected poems of Langston Hughes. New York:
Alfred A. Knopf, Inc., 1966.

 

Hutten, E. H. The origin of science. London: Allen and
Unwin, 1962.

 

Insko, C. A. & Schopler, J. Triadic consistency: A state-
ment of affective-cognitive-conative consistency.
Psychological Review, 1967, 74, 361-376.

 

Jenkins, J. J., Russell, W. A. & Suci, G. J. An atlas of
semantic profiles for 360 words. American Journal
of Psychology, 1958, 71, 688-699.

 

 

Johoda,

Jordan,

Jordan,

Jordan,

Jordan,

Jordan,

Jordan,

Jordan,

Jordan,

Kaiser,

277

M. & Warren, N. Attitudes. Baltimore: Penguin
Books, 1966.

 

J. E. Attitudes toward education and physically
disabledspersons in eleven nations. East Lansing:
Latin American Studies Center, Michigan State
University, 1968.

 

 

J. E. Attitude-behaviors toward mentally retarded
persons: A cross-cultural analysis. Final report,
U. S. Office of Education, Grant No. OEG-0-8-00126-
0197, Project No. 7—E-126, Washington, D. C., 1970.

J. E. Attitude-behavior research on physical-
mental-social disability and racial-ethnic differ-

ences. Psychological Aspects of Disability, 1971,
18, 5-26(a).

 

J. E. Construction of a Guttman facet designed
cross-cultural attitude-behavior scale toward mental
retardation. American Journal of Mental Deficiency,
1971, 76, 201-219 (b).

 

J. E. A cross-cultural facet theory paridigm for
studying attitude—behaviors toward race-ethnicity-
caste-tribalism. Abstracts of Proceedings. XX
International Congress of Psychology, August 13-19,
1972, Tokyo (a).

 

J. E. Facet theory and cross-cultural research
methodology. Paper presented at the First Inter-
national Conference for the Associa-
tion of Cross-cultural Psychology, University of
Hong Kong, August, 1972 (b).

J. E. & Bray, D. White-Black and Maori-Pakeha
attitude-behaviors in the United States and New
Zealand. Unpublished data in preparation, avail-
able from authors, 1973.

J. E. & Maierle, J. P. Guttman facet analysis of
attitudes toward mental illness, mental retardation,
and racial interaction. Paper presented at the 8th
World Congress of Mental Health, Washington, D. C.,
November 18, 1969.

H. F. Scaling a simplex. Psychometrika, 1962, 27,
155-162.

 

Kaple, J. Development of an attitude-behavior toward drug

users scale employing Guttman facet design and
analysis. Unpublished doctoral dissertation,
Michigan State University, 1971.

278

Katz, I. & Gurin, P. (Ed's.), Race and the social sciences.
New York: Basic Books, 1969.
I
Kerlinger, F. N. Foundations of behavioral research. New
York: Holt, Rinehart and Winston, 1964.

Killian, L. M. & Griggs, C. M. Racial crisis in America:
Leadership in conflict. Princeton, New Jersey:
Prentice-Hall, Inc., 1964.

 

King, M. L., Jr. The role of the behavioral scientist in
the Civil Rights Movement. In M. L. Goldschmid
(Ed.), Black Americans and white racism. New York:
Holt, Rinehart and Winston, 1970.

Kogan, N. & Downey, J. F. Scaling norm conflicts in the
area of prejudice and discrimination. Journal of
Abnormal and Social Psychology, 1956, 53, 292-295.

Kothandapani, V. Validation of feeling, belief, and inten-
tion to act as three components of attitude and
their contribution to prediction of contraceptive
behavior. Journal of Personality and Social
PsychologYI 1971, 19, 321-333.

 

Kovel, J. White racism: A psychohistory. New York:
Pantheon Bodks, 1970.

Laing, A. F. & Chazan, M. Sociometric groupings among
educationally subnormal children. American Journal
of Mental Deficiency, 1966, 71, 73-77.

 

Lange, N. New experimente fiber den vorgang der einfachen
reaktion auf sinneseindrfike. Philosophical Studies,
1888, 4, 472-510.

 

Lange, N. Beitrage sur theorie der sinnlicken aufmerksamkeit
und der activen apperception. Philosophical Studies,

1888, 4, 390-422.

Lessing, E. E. & Zagorin, S. W. Black power ideology and
college students' attitudes toward their own and
other racial groups. Journal of Personality and
Social Psychology, 1972, 21, 61-73.

Lester, J. Look out, whitey! New York: Dial Press, 1968.

 

Likert, R. A technique for the measurement of attitudes.
Archives of Psychology, 1932, 140, 1-55.

Likert, R. & Murphy, G. Public opinion and the individual.
New York: Harper, 1938.

279

Lincoln, C. E. Is anybody listening to black America?
New York: The Seabury Press, 1968.

 

Lingoes, J. C. A general survey of the Guttman-Lingoes
nonmetric program series. In R. N. Shepard, A. K.
Romney, and S. B. Nerlove (Ed's.), Multidimensional
scalin , Vol. I, New York: Seminar Press, 1972,
Pp. 49-68.

 

Lingoes, J. C. & Vandenberg, S. G. A nonmetric analysis of
twin data based on a multifaceted design. Research
report from the Louisville Twin Study Child Develop-
ment Unit, Department of Pediatrics, University of
Louisville, Report #17, April, 1966.

Maccoby, N. & Funkhouser, G. R. How do you see the city?
Psychology Today, 1968, 2, 47-50.

 

Maierle, J. P. An application of Guttman facet analysis
to attitude scale construction: A methodological
study. Unpublished doctoral dissertation, Michigan
State University, 1969.

Mann, J. H. The relationship between cognitive, affective,
and behavioral aspects of racial prejudice.
Journal of Social Psychology, 1959, 49, 223-228.

 

Mautuscelli, A. Una pesquisa sabre acrutacao de grupos
novionais, raciais e regionais un Sao Paulo.
Psicologia No. 3, Science and Letters, UfiiVersity
of Sao Paulo, 1950.

 

 

May, M. A. & Hartshorne, H. First steps toward a scale
for measuring attitudes. Journal of Educational
Psycholong 1926, 17, 145-162.

 

 

McGuire, W. J. The nature of attitudes and attitude change.
In G. Lindzey & E. Aronson (Ed's.), The handbook of
social psychology, Vol. III, Reading, Massachusetts:
Addison-Wesley, 1969, Pp. 136-314.

 

 

Mehrens, W. & Lehmann, I. J. Standardized tests in educa-
tion. New York: Holt, Rinehart and Winston, Inc.,

1968.

 

Morin, K. N. Attitudes of Texas Mexican-Americans toward
mental retardation: A Guttman facet analysis.
Unpublished doctoral dissertation, Michigan State
University, 1969.

Morland, J. K. & Williams, J. E. Cross-cultural measurement
of racial and ethnic attitudes by the semantic
differential. Social Forces, 1969, 48, 107-113.

 

280

Mukherjee, B. N. Deviation of likelihood-ratio tests for
Guttman quasi-simplex covariance structures.
Psychometrika, 1966, 31, 97-123.

 

Munsterberg, H. Beitrage sur expsrimentellen_psychologie.
Vol. I. Freiburg: Morh, 1889.

 

Murray, H. A. Explorations in personality. New York:
Oxford University Press, 1947.

 

Newman, J. R. The world of mathematics. Vol. III, New
York: Simon and Schuster, 1956.

 

Nicholson, W. N. A Guttman facet analysis of attitude-
behaviors toward drug users by heroin addicts and
mental health therapists. Unpublished doctoral
dissertation, Michigan State University, 1972.

Osgood, C. E. Cross-cultural comparability in attitude
measurement via multilingual semantic differentials.
In I. D. Steiner & M. Fishbein (Ed's.), Current
studies in social psychology, New York: Holt,
Rinehart and Winston, 1965, Pp. 95-107.

 

Osgood, C.E. Probing subjective culture. Chapter 6 of
book in process, mimeo, 1973 (scheduled).

Osgood, C. E., Suci, G. J. & Tannenbaum, P. H. The measurement
of meaning. Illinois: University of Illinois Press,
I957.

 

 

Ostrom, T. M. The emergence of attitude theory: 1930-
1950. In A. G. Greenwald, T. C. Brock & T. M.
Ostrom (Ed's.), Psychological foundations of
attitudes. New York: Academic Press, 1968.

 

 

Ostrom, T. M. Item construction in attitude measurement.
Public Qpinion Quarterly, 1971-1972, 35, 593-600.

 

Peters, R. S. (Ed.). Brett's history of psychology.
London: Allen 3 Unwin, 1962.

 

Peterson, H. A. Correlation of certain mental traits in
normal school students. Psychological Review,
1908, 15, 323-338.

 

Pettigrew, T. F. A profile of the Negro American. Prince-
ton, New Jersey: D. Van Nostrand, 1964.

 

Podair, S. Language and prejudice toward Negroes. In
B. N. Schwartz and R. Disch (Ed's.), White racism:
Its' history, pathology and practice. New York:
Dell Publishing Company, 1970, Pp. 385-391.

 

 

281

Poulos, T. H. Attitudes toward the deaf: A Guttman facet
analysis of their content, structure and determi-
nants. Unpublished doctoral dissertation, Michigan
State University, 1970.

Proshansky, H. & Newton, P. The nature and meaning of
Negro self-identity. In M. Deutsch, I. Katz, and
A. R. Jensen (Ed's.), Social class, race, and
psychological development. New York: Holt,
Rinehart and Winston, 1968, Pp. 178-218.

 

 

Proenza, L. & Strickland, B. R. A study of prejudice in
Negro and White college students. Journal of
Social Psychology, 1965, 67, 273-281.

 

 

Renninger, C. A. & Williams, J. E. Black-white color
connotations and racial awareness in preschool
children. In M. L. Goldschmid (Ed.), Black Ameri-
cans and white racism. New York: Holt, Rinehart
and Winston, 1970, Pp. 311-320.

 

Report of the national advisory commission on civil dis-
orders. New York: Bantom Books, 1968.

Rokeach, M. Beliefs, attitudes and values. San Francisco:
Jossey-Bass, I968.

 

Rosander, A. C. An attitude scale based upon behavior
situations. Journal of Social Psyshology, 1937,
8, 3-16.

 

Rosenberg, M. J. & Hovland, C. I. Cognitive, affective,
and behavioral components of attitudes. In M. J.
Rosenberg & C. I. Hovland (Ed's.), Attitude orgsn-
ization and change. Vol. III, New Haven & London:
Yale University Press, 1960, Pp. 1-14.

 

Rotter, J. B. Generalized expectancies for internal versus
external control of reinforcement. Psychological
Monogrsphs, 1966, 80, 1, #609, whole issue.

 

Rotter, J. B., Seeman, M. & Liverant, S. Internal versus
external control of reinforcements: A major varia-
ble in behavior therapy. In N. F. Washburne (Ed.),
Decisions, values and groups, Vol. 2, London:
Pergamon, 1962, Pp. 473-516.

Ruble, W. L., Kiel, D. F. & Rafter, M. E. Analysis of
covariance and analysis of variance with unequal
frequencies permitted in the cell TLS routine).
Stat. Series Description No. 115, Agricultural
Experiment Station, Michigan State University, 1966.

282

Schwartz, B. N. & Disch, R. (Ed's.). White racism: Its'
history, psthology and practice. New York: Dell
Publishing Company, 1970.

Schonnell, F. S. & Rorke, M. A second survey of the effects
of a subnormal child on the family unit. American
Journal of Mental Deficiency, 1959163! 566-574-

Schonnell, F. S. & Watts, F. F. A first survey of the
effects of a subnormal child on the family unit.
American Journal of Mental Deficiency, 1956, 61,
210-219.

Scott, W. A. Attitude measurement. In G. Lindzey & E.
Aronson (Ed's.), The handbook of social psychology.
Reading, Massachusetts: Addison-Wesley, 1969,

Vol. II, Pp. 204-273.

Seward, G. Psychotherapy and culture conflict. New York:
Ronald Press, 1956.

Shaw, M. E. & Wright, J. M. Scales for the measurement of
attitudes. New York: McGraw—Hill, 1967.

 

Sikes, M. P. The counseling psychology curriculum: A)
new dimension. The Counseling Psychologist, 1971,
2, 102-104.

Smith, W. A facet analysis of racial attitudes in the
United States and Africa. Unpublished doctoral
dissertation, Michigan State University, 1973
(scheduled).

Smith, W. G. Return to Black America. New Jersey:
Prentice-Hail, Inc., 1970.

Spearman, C. The abilities of man. New York: MacMillan,
1927.

Spearman, C. & Jones, L. W. Human ability. London:
MacMillan, 1951.

Stace, W. T. A critical history of Greek philosophy.
London: MacMillan, 1967.

Stickler, G. Authoritarian ideology in Negro college
students. Journal of Abnormal and Social Psycholpgy,

1957, 54, 396-399.

Stockton, J. L. The definition of intelligence in relation
to modern methods of mental measurements. Psycho-
logical Monpgrsphs, 30, 1921, #137, whole issue.

283

Sully, J. The human mind. New York: D. Appleton, Vols.
I & II, 1892.

 

Szwed, J. F. Discovering Afro-America. In J. F. Szwed
(Ed.), Black America. New York: Basic Books,
1970, Pp. 286-296.

 

Taylor, T. P. A study of the relationship between demo-
graphic variables and racial attitudes. Unpublished
doctoral dissertation, Michigan State University,
1971.

Thomas, W. I. & Zananiechi, F. The Polish peasant in Europe
and America. Boston: Badger, 1918-1920.

 

 

Thurstone, L. L. The vectors of the mind. Chicago:
University of Chicago Press, 1935.

 

Thurstone, L. L. Multiple-factor analysis. Chicago:
University of Chicago Press, 1947.

 

Triandis, H. C. An analysis of supjective culture. New
York: Wiley, 1972.

 

Triandis, H. C., Loh, W. D. & Levin, A. Race, status,
quality of spoken English, & opinions about civil
rights as determinants of interpersonal attitudes.

Journal of Personality and Social Psychology,
1966, 3, 468-472.

 

Triandis, H. C., Malpass, R. S. & Davidson, A. R. Cross-
cultural psychology. The Biennial Review of
Anthrspology, 1972, 1-84.

 

 

United States Commission of Civil Rights. Racial isolation
and the outcomes of education. In M. L. Goldschmid
(Ed.), Black Americans and white racism. New
York: Holt, Rinehart and Winston, 1970, Pp. 226-
252.

 

Valentine, C. A. Deficit, difference and bicultural models
of Afro-American behavior. Harvard Educational
Review, 1971, 41, 137-157.

 

Vinacke, W. E. Explorations in the dynamic processes of
stereotyping. Journal of Social Psychology, 1956,
43, 104-132.

 

Vurdelja, D. Attitudes of mothers of retarded and non-
retarded in four nations: A Guttman facet analysis.
Unpublished M. A. thesis, Michigan State University,
1970.

284

Ward, J. Psychological principles. Cambridge, Massachu-
setts: University Press, 1920.

 

Webster, S. W. The influence of interracial contact on
social acceptance in a newly integrated school.
Journal of Educational Psychology, 1961, 52, 292-
296.

 

Weiss, W. An examination of attitudes toward Negroes.
Journal of Social Psychology, 1961, 15, 3-31.

 

Williams, J. E. Connotations of color names among Negroes

and Caucasians. Perceptual and Motor Skills, 1964,
18, 721-731.

 

Williams, J. E. Connotations of racial concepts and color
names. Journal of Personality and Social Psycho-

 

Williams, J. E. Individual differences in color-name
connotations as related to measures of racial
attitude. Perceptual and Motor Skills, 1969, 29,
383-386.

 

Williams, J. E. Connotations of racial concepts and color
names. In M. L. Goldschmid (Ed.), Black Americans
and white rscism. New York: Holt, Rinehart and
Winston, 1970, 38-48.

 

 

Williams, J. E., Best, D. L., Wood, F. B. and Filler, J. W.
Changes in the connotations of racial concepts and
color names: 1963-1970, in process, 1971, mimeo.

Williams, J. E. & Carter, D. J. Connotations of racial
concepts and color names in Germany. Journal of
Social Psychology, 1967, 72, 19-26.

 

 

Williams, J. E. & Edwards, C. D. An exploratory study of
the modification of color and racial concept atti-
tudes in preschool children. Child Development,
1969, 40, 737-750.

 

Williams, J. E., Morland, J. K. & Underwood, W. L.
Connotations of color names in the United States,
Europe, and Asia. Journal of Social Psychology,
1970, 82, 3-14.

 

Williams, J. E. & McMurtry, C. A. Color connotations among
Caucasian seventh graders and college students.
Percsptual and Motor Skills, 1970, 30, 707-713.

 

285

Williams, J. E. & Roberson, J. K. A method for assessing

racial attitudes in preschool children. Educa-

tional and Psychological Measurement, 1967, 27,
671-689.

Williams, J. E., Tucker, R. D. & Dunham, F. Y. Changes in
the connotations of color names among Negroes and
Caucasians: 1963-1969. Journal of Personality
and Social Psychology: 1971, 19, 222-229.

Williams, R. L. The changing image of the black American:
A socio-psychological appraisal. In C. Thomas

(Ed.), Boys no more: A black psychologist's view
of community. Beverly Hills, California: Glencoe
Press, 1972, Pp. 66-79.

Williams, R. L. & Kirkland, J. The White Counselor and the
Black client. The Counseling Psychologist, 1971,
2, 114-117.

Williams, W. S. Attitudes of black and white policemen

toward the opposite race. Unpublished doctoral
dissertation, Michigan State University, 1970.

Wolf, R. M. Construction of descriptive and attitude

scales. In T. Husen (Ed.), Intespational stusy of
achievement in mathematics, Vol. I. New York:
John Wiley and Sons, 1967.

Yuker, H. E. Attitudes as determinants of behavior.

Journal of Rehabilitation, 1965, 31, 15-16.

Zeller, E. A history of Greek philosophy. Vol. II,
London:

Longmans, Green, and Company, 1881.

APPENDICES

286

APPENDIX A

GLOSSARY

287

GLOSSARYl

Approximation--see "simplex approximation".

Attitude--"De1imited totality of behavior with respect to something"
(Guttman, 1950a, p. 51)

Attitude-behavior--the hyphenated term denotes that attitude is a
subclass of behavior rather than an intervening variable or
a "predisposition" to behavior.

Content--situation (action, feeling, comparison, circumstances)
indicated in an attitude item; generally corresponds to
"lateral struction".

Definitional statement--specification of characteristics proper to
an item of a given Level member, typically stated in phrase
or clause form.

Definitional system-~ordered group of definitional statements or
of the corresponding Level members; typically either the
group constituting a "semantic path” or the complete group
of 12 Level members in the "semantic map".

Directionality--characteristic of an item, sometimes called positive
or negative, determining agreement with the item as
indicating favorableness or unfavorableness toward the
attitude object.

Element--one of two or more ways in which a facet may be expressed;
in the present system, all joint facets are dichotomous,
expressed in one of two ordered elements.

Facet--one of several semantic units distinguishable in the verbal
expression of an attitude; in the present system, five

dichotomous facets are noted within the joint struction.

 

1Credit is given to Maierle (1969) for most of the work in
developing this glossary.

288

289

Facet profile--see "struction profile".

Joint struction-~see also "struction", "lateral struction"--
"operationally defined as the ordered sets of . . . five
facets from low to high across all five facets simul-
taneously” (Jordan, 1968, p. 76); that part of the semantic
structure of attitude items which can be determined
independently of specific response situations.

Lateral struction--see also "struction", "joint struction"--that
part of the semantic structure of attitude items which is
directly dependent on specification of situation and object;
a more precise term than "content”.

Level-~degree of attitude strength specified by the number of
strong and weak facets in the member(s) of that Level; in
the present system, six ordered Levels are identified:
Level 1 is characterized by the unique member having five
weak facets; Level 2, by members having four weak and one
strong facet . . . Level 6, by the unique member having five
strong facets.

Level member-~one of one or more permutation(s) of strong and weak
facets which are common to a given Level; in the present
system, 12 Level members have been identified: three on
Level 2, four on Level 3, two on Level 4, and one each on
Levels 1, 5, and 6.

Map-~see "semantic map".

Member--see "Level member".

Path-~see "semantic path".

Profile--see "struction profile".

290

Reversa1--change in a specified order of Levels or of correlations,
involving only the two indicated Levels or correlations.

Semantic--pertaining to or arising from the varying meanings,
grammatical forms, or stylistic emphasis of words, phrases,
or clauses.

Semantic map--two-dimensiona1 representation of hypothesized
relationships among six Levels and among 12 Level members.

Semantic path--ordered set of Level members, typically six, such
that each member has one more strong facet than the
immediately preceding member and one less strong facet then
the immediately following member.

Semantic possibility analysis--1inguistic discussion of the
implications of the five dichotomous joint facets identified
in the present system; of 32 permutations, only 12 are
considered logically consistent.

Simplex--specific form of (correlation) matrix, diagonally dominated
and decreasing in magnitude away from the main diagonal.

Simplex approximation--matrix which approaches more or less perfectly
the simplex form; existing tests (Kaiser, 1962; Mukherjee,
1966) reflect both ordering of individual entries and sizes
of differences between entries and between diagonals.

Strong(er)--opposite of weak(er)--term functionally assigned to one
of two elements, to a facet expressed by its strong element,
or to a Level member characterized by more strong facets
than another Level member; the strong-weak continuum is

presently examined as unidimensional.

291

Struction—-see also ”joint struction", ”lateral struction”--semantic
pattern identifiable in any attitude item, or the system of
such identifications.

Struction profile--specification, typically indicated by small
letters and numerical subscripts, of the permutation(s)
of weak and strong elements or facets in a Level member
or a set of Level members; or of permutations of lateral
elements or facets.

Transposition--change in a specified order of Levels or of
correlations involving a change in position of one Level
or correlation and the corresponding one-place shift in
the position of following or preceding Levels or correlations.

Weak--opposite of ”strong" (which see).

APPENDIX B-l

ATTITUDE BEHAVIOR SCALE WB-G

292

ATTITUDE BEHAVIOR SCALE-WB-G

DIRECTIONS

This booklet contains statements of how people behave in certain situations
or feel about certain things. You, yourself, or other White persons often
behave in the same way toward Blacks. You also have some general ideas
about yourself, about other White persons like you and about Blacks. Some-
times you feel or behave the same way toward everyone and sometimes you
feel or behave differently toward Blacks.

 

This questionnaire has statements about ideas and about behavior. Each
statement of this questionnaire is different from every other section,
although some of the statements in eaCh section are similar. Your answers
in one section, therefore, may be the same as answers in another section,
or your answers may differ; from section to section. Here is a sample
statement:

Sample'l
Other Whites believe the following things about Blacks as compared to Whites:
1. Chance of Blacks being taller

less chance than Whites
. about the same
3. more chance than Whites

If other Whites believe that Blacks have less chance than Whites to be
taller, you should circle the number 1 as shown above or if you are using
an IBM answer sheet make a heavy dark line on the answer sheet between the
two lines after the number as follows:

 

1. 1 — 2 --- 3 --- 4 --- 5 H-..

*************** DO NOT PUT YOUR NAME ON THE BOOKLET ***************

©

Richard J. Hamersma

John E. Jordan

112272 College of Education
Michigan State University

by:

-1-

ABS-I-WB-G

 

Directions: Section I

This section contains statements about ideas which other Whites have about
Blacks. Circle 25 fill ip_the.answer sheet number that indicates how
other Whites compare Blacks with themselves. Please answer all questions.

 

Other Whites believe the following things about Blacks as compared to Whites:

 

l. Blacks can be trusted with money

1. less than Whites
2. about the same as Whites
3. more than Whites

2. Black families are closely knit
1. less often than White ones

2. about as often as White ones
3. more often than White ones

3. Blacks' intellectual ability is
1. less than Whites'
2. about the same as Whites'
3. more than Whites'

4. Blacks desire a higher education
1. less often than Whites
2. about as often as Whites
3. more often than Whites

5. Blacks help their neighbors

1. less than Whites do
2. about the same as Whites do
3. more than Whites do

6. Black neighborhoods are safe
1. less often than White ones
2. about as often as White ones
3. more often than White ones

7. Blacks obey job rules and regulations

1. less than Whites
2. about the same as Whites
3. more than Whites

8. 'Whites enjoy working with Blacks
1. less than Blacks do with Whites

2. about the same as Blacks
3. more than Blacks do with Whites

-2-

ABS-I-WB-G

 

Other Whites believe the following things about Blacks as compared

 

to Whites:

9.

10.

ll.

12.

13.

14.

112272

Blacks resist arrest

1. more than Whites
2. about the same as Whites
3. less than Whites

Blacks are victims of "police brutality"
l. more than Whites

2. about the same as Whites

3. less than Whites

Blacks misuse trial-by-jury

l. more often than Whites

2. about as often as Whites

3. less often than Whites

Whites vote for Black candidates for public office
1. less than Blacks do for Whites

2. about the same as Blacks do for Whites
3. more than Blacks do for Whites

Blacks desire draft deferments

l. more often than Whites

2. about as often as Whites

3. less often as Whites

Blacks are careful with their weapons

1. less often than Whites

2. about as often as Whites
3. more often than Whites

112272

-3-

ABS-II-WB-G

Directions: Section II

 

This section contains statements about things which most Whites generally
believe others would experience when interacting with Blacks. Please choose
the answer that indicates what you think most others believe about Blacks.

 

Most Whites generally believe the following about interacting with Blacks:

 

15.

16.

17.

l8.

19.

20.

21.

Whites believe they can trust Blacks with money

1. disagree
2. uncertain
3. agree

Whites believe that Black families are as closely knit as their own
1. disagree

2. uncertain

3. agree

Whites believetfluaintellectual ability of Blacks is equal to theirs
1. disagree

2. uncertain

3. agree

Whites believe they want to do their study or school work with Blacks
1. disagree

2. uncertain

3. agree

Blacks like to help their neighbors

1. less than Whites do

2. abou t the same

3. more than Whites do

Whites believe that Black neighborhoods are safe for Whites

1. disagree

2. uncertain

3. agree

Whites believe Blacks obey job rules and regulations as much as they do
1. disagree

2. uncertain
3. agree

112272

-4-

ABS-II-WB-G

 

Most Whites generally believe the following about interacting with Blacks:

 

22.

23.

24.

25.

26.

27.

28.

Whites believe they enjoy working with Blacks

1. disagree
2. uncertain
3. agree

Blacks believe in resisting arrest from White policemen

1. agree
2. uncertain
3. disagree

Whites believe that Blacks are the victims of "police brutality"
from Whites

1. agree
2. uncertain
3. disagree

Whites usually use trial-by-jury fairly when they deal with Blacks
1. usually not believed in

2. undecided

3. usually believed in

For Whites to vote with Blacks for Black candidates for public office
1. usually not approved

2. undecided

3. usually approved

For Whites to be given draft deferments more than Blacks

1. usually approved

2. undecided

3. usually not approved

For Blacks to be as careful with their weapons as Whites are

1. usually not approved

2. undecided
3. usually approved

 

 

 

q"

112272

-5-

ABS-III-WB-G

Directions: Section III

 

This section contains statements of the right or wrong way of behaving or
acting toward Blacks. You are asked to indicate what you yourself believe
others think should be done with respect to Blacks.

In respect to Blacks, what do you believe others think is right

 

or wrong?
29. To trust Blacks with money is
1. usually wrong
2. 'undecided
3. usually right
30. To expect Black families to be as closely knit as White ones is
1. usually wrong
2. undecided
3. usually right
31. To expect Blacks' intellectual ability to be the same as Whites is
1. usually wrong
2. undecided
3. usually right
32. To expect Blacks to desire a higher education as much as Whites do is
1. usually wrong
2. undecided
3. usually right
33. To expect Whites to help Black neighbors is
1. usually wrong
2. undecided
3. usually right
34. To expect Whites to believe that Black neighborhoods are safe for them is
1. usually wrong
2. undecided
3. usually right
35. To expect Blacks to obey job rules and regulations the same as Whites do is
1. usually wrong
2. undecided
3. usually right

12272

-6-

ABS-III-WB-G

 

In respect to Blacks, what do you believe others think is right or wrong:

36.

37.

38.

39.

40.

41.

42.

To expect Whites to

1. usually wrong
2. undecided
3. usually right

To expect Blacks to

1. usually right
2. undecided
3. usually wrong

To expect Blacks to

1. usually right
2. undecided
3. usually wrong

To expect Whites to

1. usually right
2. undecided
3. usually wrong

To expect Whites to
office is

1. usually wrong
2. undecided
3. usually right

To expect Blacks to

1. usually wrong

2. undecided

3. usually right

To expect Blacks to
. usually wrong

1
2. undecided
3. usually right

enjoy working with Blacks is

resist arrest from White officials is

be the victims of "police brutality" from Whites is

misuse trial-by-jury when they deal with Blacks is

vote with Blacks for Black candidates for public

be given draft deferments equally with Whites is

be as careful with their weapons as Whites are is

112272

-7-

ABS-IV-WB-G

 

Directions: Section IV

This section contains statements about how you think you would act toward
Blacks. Choose the answer that indicates how you think you would act.

 

In respect to a Black person would ypg yourself:

43.

44.

45.

46.

47.

48.

49.

I would trust Blacks with money

1. no
2. undecided
3. yes

I would want my family to be as closely knit as Black families are

1. no
2. undecided
3. yes

I would want my intellectual ability to be the same as that of Blacks

1. no
2. undecided
3. yes

I would want to have the same desire Blacks do for a higher education

1. no
2. undecided
3. yes

I would help Black neighbors

I. no
2. undecided
3. yes

I would want White neighborhoods to be as safe as I think Black ones are

1. no
2. undecided
3. yes

I would obey job rules and regulations the same as Blacks

1. no
2. undecided
3. yes

112272

-3-

ABS-IV-WB-G

In respect to a Black person would you yourself:

50.

51.

52.

53.

54.

55.

56.

I would enjoy working with Blacks

1. no
2. undecided
3. yes

I would resist arrest if arrested by Blacks

1. yes
2. undecided
3. no

I would expect "police brutality" from Blacks

1. yes
2. undecided
3. no

WOuld you use trial-by-jury equally when dealing with Blacks?

1. no
2. undecided
3. yes

Would you vote for a Black candidate for public office?

1. no
2. undecided
3. yes

WOuld you want Whites to be given draft deferments as much as Blacks?

1. no
2. undecided
3. yes

WOuld you be as careful with weapons as you think Blacks are?

1. no
2. undecided
3. yes

112272

-9-

ABS-V-WB-G

Directions: Section V

This section concerns actual feelings that you yourself may have about Blacks.
You are asked to indicate how you feel about the following statements.

 

 

How do you actually feel toward Blacks:

57. When Whites trust Blacks with money I feel

1. bad
2. indifferent
3. good
58. When White families are as closely knit as I think Black families are
I feel ‘
1. bad
2. indifferent
3. good

59. When Blacks' intellectual ability is the same as Whites I feel

1. bad
2. indifferent
3. happy

60. When Blacks desire a higher education as much as Whites do, I feel

1. bad
2. indifferent
3. good

61. When Whites help Black neighbors I feel

1. bad
2. indifferent
3. good

62. When Whites are safe in Black neighborhoods, I feel

1. bad
2. indifferent
3. good

63. When Blacks obey job rules and regulations the same as Whites, I feel

1. dissatisfied
2. indifferent
3. satisfied

-10-

ABS-V-WB-G

How do you actually feel toward Blacks:

64.

65.

66.

67.

68.

69.

70.

112272

When Whites enjoy working with Blacks I feel

1. bad
2. indifferent
3. good

When Blacks resist arrest the same as Whites, I feel
1. dissatisfied

2. indifferent

3. satisfied

When Blacks use "police brutality" the same as Whites do, I feel

1. bad
2. indifferent
3. good

When Whitesrmisuse trial-by-jury in relation to Blacks, I feel

1. happy
2. indifferent
3. angry

When Whites vote for Black candidates for public office, I feel

1. bad
2. indifferent
3. good

When Blacks are given draft deferments as much as Whites, I feel

1. dissatisfied
2. indifferent
3. satisfied

When Blacks are as careful with their weapons as Whites are, I feel

1. dissatisfied
2. indifferent
3. satisfied

112272

-11-

ABS-VI-WB-G

Directions: Section VI

This section concerns actual experiences you have had with Blacks. Try to
answer the following questions from the knowledge of your own experience. If
you have had pg experience or contact with Blacks, omit questions 71-84 and
begin again at question number 85. If you have had any experience or contact
with Blacks answer all of the following questions. -_-

 

 

Experiences or contacts with Blacks:

 

71.

72.

73.

74.

75.

76.

77.

I have trusted Blacks with money

1. no
2. uncertain 3
3. yes

I have seen that Black families are as closely knit as White ones

1. no
2. uncertain
3. yes

The intellectual ability of Blacks is equal to mine

1. no
2. uncertain
3. yes

The Blacks I know wanted a higher education as much as I did

1. no
2. uncertain
3. yes

I have helped a Black neighbor

1. no
2. uncertain
3. yes

I have felt safe when in Black neighborhoods

1. no
2. uncertain
3. yes

I have seen that Whites obey job rules and regulations when working
with Blacks '

1. no
2. uncertain
3. yes

112272

-12-

ABS-VI-WB-G

 

Experiences or contacts with Blacks:

 

78.

79.

80.

81.

82.

83.

84.

I have enjoyed working with Blacks

1. no
2. uncertain
3. yes

I have resisted arrest by Blacks

1. yes
2. uncertain
3. no

I have been the victim of "police brutality" from Blacks

1. yes
2. uncertain
3. no

I have seen Whites misuse trial-by-jury with Blacks

1. yes
2. uncertain
3. no

I have seen that Whites vote for Black candidates for public office

1. no
2. uncertain
3. yes

I have seen that Blacks are given draft deferments as much as Whites

1. no
2. uncertain
3. yes

I have seen that Blacks are as careful with their weapons as Whites

1. no
2. uncertain
3. yes

-13- ABS-WB-D

ATTITUDE BEHAVIOR SCALE - ABS-WB-D

This part of the questionnaire deals with many things. For the purpose of

this study, the answers of all persons are ipportant.

 

 

Part of the qeustionnaire has to do with personal information about you.
Since the questionnaire is completely anonymous or confidential, you may
answer all of the questions freely without any concern about being identified.

l£_is ipportant £o_the study £p_obtain your answer go every gpestion.

Please read each question carefully and so not omit any questions. Please

answer by circling the answer or marking the space on the IBM answer sheet.

85. Please indicate your sex.
1. Female
2. Male

86. Please indicate your age as follows:

1. Under 20
2. 21-30
3. 31-40
4. 41-50

5. Sl-over
87. What is your marital status?
1. Married
2. Single
3. Divorced
4. Widowed

S. Separated

112272

[12272

88.

89.

90.

91.

-l4- ABS-WB-D

What is your religion?

1. I prefer not to answer

2. Catholic

3. Protestant

4. Jewish

5. Other

Please indicate level of education

1. First year university

2. Second year university

3. Third year university

4. Fourth year university

5. Graduate student

Some people feel that in bringing up children, new ways and methods should
be tried whenever possible. Others feel that trying out new methods is
dangerous. What is your feeling about the following statement?

"New methods of raising children should be tried whenever possible."

1. Strongly disagree

2. Slightly disagree

3. Slightly agree

4. Strongly agree

Family planning on birth control has been discussed by many people. What
is your feeling about a married couple practicing birth control?

1. It is always wrong
2. It is usually wrong
3. It is probably all right

4. It is always right

112272

92.

93.

94.

95.

-15- ABS-WB-D

The following questions have to do with kinds of experiences you have
had with Blacks. If more than one experience applies, please choose the
answer with the highest number.

 

 

l. I have read or studied about Blacks through reading, movies, lecture
or observation.

2. A friend or relative is a Black person.

3. I have personally worked with Blacks as teacher, counselor, volunteer,
child care, etc.

Considering all of the times you have talked, worked or in some other
way had personal contact with Blacks about how much has it been altogether?

I. Only a few casual contacts.

2. Between one and three months.

3. Between three and six months.

4. Between six months and one year.
5. More than one year of contact.

When you have been in contact with Blacks, how easy for you, in general,
would you say it would have been to have avoided being with them?

1. I have had no contact.

2. I could generally have avoided these personal contacts only at great
cost or difficulty.

3. I could generally have avoided these personal contacts only with
considerable difficulty.

 

4. I could generally have avoided these personal contacts but with some
inconvenience.

5. I could generally have avoided these personal contacts without any
difficulty or inconvenience.

 

If you have ever worked with Blacks for personal gain (for example, for
money or some other gain) what opportunities did you have (or do you have)
to work at something else instead; that is, something else that was (is)
acceptable to you as a job?

1. No such experience.

2. No other job available.

3. Other jobs available not s5 all acceptable to me.

4. Other jobs available were not guite acceptable to me.

5. Other jobs available were fully acceptable to me.

112272

96.

97.

98.

99.

-l6- ABS-WB-D

How have you generally felt about your experiences with Blacks?

no experience.

I definitely dislike it.

I did not like it very much.
I like it somewhat.

I definitely enjoyed it.

Which of the following do you think would have the greatest effect of
reducing racial prejudice? Circle only one or mark only one on the IBM
answer sheet.

1. Integration of schools.

2. Publicity campaigns to promote integration.

3. Fair employment legislation.

4. Open housing legislation.

5. Direct, personal contact between members of various racial groups.
How would you rate your own racial attitudes as compared to the average
person?

1. Very much more prejudiced.

2. Somewhat more prejudiced.

3. About the same.

4. Somewhat less prejudiced.

5. Very much less prejudiced.

To which racial group do you belong?

Prefer not to answer.

White

Black

Oriental

Other

-17- ABS-WB-D

100. Where were you mainly reared or "brought up":h1youryouth (that is, up
to age 21)?

1. country
2. country town
3. city suburb

4. city

112272

112272

-l8- ABS-WB-D

LIFE SITUATIONS

This section of the booklet deals with how people feel about several aspects of
life or life situations. Please indicate how you feel about each situation by
circling the answer you choose or marking on the IBM answer sheet.

101.

102.

103.

104.

105.

106.

It

waI—I
.0.

should be possible to eliminate war once and for all

strongly disagree

disagree
agree

strongly agree

Success depends to a large part on luck and fate

waI-t
.0.

strongly
agree

disagree
strongly

Someday most

but)?!
0..

By

42‘me

strongly
disagree
agree

strongly

improving

strongly
disagree
agree

strongly

agree

disagree
of the mysteries of the world will be revealed by science

disagree

agree
industrial and agricultural methods, poverty can be eliminated

disagree

agree

With increased medical knowledge it should be possible to lengthen the
average life span to 100 years or more

$~u>n>ha

strongly
disagree
agree

strongly

disagree

agree

Someday the deserts will be converted into good farming land by the
application of engineering and science

£~uahara
0

strongly
disagree
agree

strongly

disagree

agree

-19- ABS-WB-D

107. Education can only help people develop their natural abilities; it
cannot change peOple in any fundamental way.

strongly agree
agree

disagree

strongly disagree

¢~uan>e:
. O O

108. With hard work anyone can succeed

I. strongly disagree
2. disagree

3. agree

4

strongly agree

109. Almost every present human problem will be solved in the future

strongly disagree
disagree

agree

strongly agree

waI—I
.00

112272

APPENDIX B-2

ATTITUDE BEHAVIOR SCALE WN-G

313

112272

ATTITUDE BEHAVIOR SCALE-WN-G

DIRECTIONS

 

This booklet contains statements of how people behave in certain situations
or feel about certain things. You, yourself, or other White psrsons often
behave in the same way toward Negroes. You also have some general ideas
about yourself, about other White persons like you and about Negroes.
Sometimes you feel or behave the same way toward everyone and sometimes you
feel or behave differently toward Negroes.

 

This questionnaire has statements about ideas and about behavior. Each
statement of this questionnaire is different from every other section,
although some of the statements in each section are similar. Your answers
in one section, therefore, may be the same as answers in another section,
or your answers may differ from section to section. Here is a sample
statement:

Sample 1

Other Whites believe the following things about Negroes as compared to Whites:
1. Chance of Negroes being taller

<;> less chance than Whites
. about the same
3. more chance than Whites

If other Whites believe that Negroes have less chance than Whites to be
taller, you should circle the number 1 as shown above or if you are using
an IBM answer sheet make a heavy dark line on the answer sheet between the
two lines after the number as follows:

 

 

 

1. 1 ’ 2 === 3 =:: 4 ::: 5 ==:

***************4 DO NOT PUT YOUR NAME ON THE BOOKLET *****************

by: (Ekichard J. Hamersma
John E. Jordan
College of Education
Michigan State University

ABS-I-WN-G

 

Directions: Section I

 

This section contains statements about ideas which other Whites have about
Negroes. Circle or fill in the answer sheet number that indicates how
other Whites compare Negroes with themselves. Please answer all questions.

 

Other Whites believe the following things about Negroes as compared to Whites:

 

l. Negroes can be trusted with money

1. less than Whites
2. about the same as Whites
3. more than Whites

2. Negro families are closely knit

1. less often than White ones
2. about as often as White ones
3. more often than White ones

3, Negroes' intellectual ability is

1. less than Whites'
2. about the same as Whites'
3. more than Whites'

4. Negroes desire a higher education

1. less often than Whites
2. about as often as Whites
3. more often than Whites

5. Negroes help their neighbors

1. less than Whites do
2. about the same as Whites do
3. more than Whites do

6. Negro neighborhoods are safe

1. less often than White ones
2. about as often as White ones
3. more often than White ones

7. Negroes obey job rules and regulations

1. less than Whites
2. about the same as Whites
3. more than Whites

8. Whites enjoy working with Negroes

1. less than Negroes do with Whites
2. about the same as Negroes

“2272 3. more than Negroes do with Whites

112272

-2-

ABS-I-WN-G

 

Other Whites believe the following things about Negroes as compared to

 

Whites:

9.

10.

ll.

12.

l3.

l4.

Negroes resist arrest

1. more than Whites
2. about the same as Whites
3. less than Whites

Negroes are victims of "police brutality"
l. more than Whites

2. about the same as Whites

3. less than Whites

Negroes misuse trial-by-jury

l. more often than Whites

2. about as often as Whites
3. less often than Whites

Whites vote for Negro candidates for public office
1. less than Negroes do for Whites

2. about the same as Negroes do for Whites

3. more than Negroes do for Whites

Negroes desire draft deferments

l. more often than Whites
2. about as often as Whites
3. less often than Whites

Negroes are careful with their weapons
1. less often than Whites

2. about as often as Whites
3. more often than Whites

11272

-J-

ABS-II-WN-G

 

Directions: Section II

 

This section contains statements about things which most Whites generally
believe others would experience when interacting with Negroes. Please choose
the answer that indicates what you think most others believe about Negroes.

 

 

 

Most Whites generally believe the following about interacting with Negroes:

15.

l6.

l7.

l8.

19.

20.

21.

Whites believe they can trust Negroes with money

1. disagree
2. uncertain
3. agree

Whites believe that Negro families are as closely knit as their own
1. disagree

2. uncertain
3. agree

Whites believe the intellectual ability of Negroes is equal to theirs
1. disagree

2. uncertain

3. agree

Whites believe they want to do their study or school work with Negroes
1. disagree

2. uncertain

3. agree

Negroes like to help their neighbors

1. less than Whites do

2. about the same
3. more than Whites do

Whites believe that Negro neighborhoods are safe for Whites

1. disagree

2. uncertain

3. agree

Whites believe Negroes obey job rules and regulations as much as they do
1. disagree

2. uncertain
3. agree

-4-

ABS-II-WN-G

Most Whites generally believe the following about interacting with Negnoes:

22.

23.

24.

25.

26.

27.

28.

112272

Whites believe they enjoy working with Negores

1. disagree
2. uncertain
3. agree

Negroes believe in resisting arrest from White policemen

1. agree
2. uncertain
3. disagree

Whites believe that Negroes are the victims of "police brutality"
from Whites

1. agree
2. uncertain
3. disagree

Whites usually use trial-by-jury fairly when they deal with Negroes
1. usually not believed in
2. undecided
3. usually believed in
For Whites to vote with Negroes for Negro candidates for public office
1. usually not approved
2. undecided
3. usually approved
For Whites to be given draft deferements more than Negroes
1. usually approved
2. undecided
3. usually not approved
For Negroes to be as careful with their weapons as Whites are
. usually not approved

1
2. undecided
3. usually approved

[2272

‘J"

ABS-III-WN-G

 

Directions: Section III

 

This section contains statements of the right or wrong way of behaving or acting
toward Negroes. You are asked to indicate what you yourself believe others
think should be done with respect to Negroes.

 

In respect to Negroes, what do you belieye others think is right or wrong:

29.

30.

31.

32.

33.

34.

35.

To

1.
20
3

trust Negroes with money is

usually wrong
undecided
usually right

expect Negro families to be as closely knit as White ones is
usually wrong

undecided

usually right

expect Negroes' intellectual ability to be the same as Whites is
usually wrong

undecided

usually right
expect Negroes to desire a higher education as much as Whites do is
usually wrong

undecided

usually right

expect Whites to help Negro neighbors is

usually wrong

undecided

usually right

expect Whites to believe that Negro neighborhoods are safe for them is
usually wrong

undecided

usually right

expect Negroes to obey job rules and regulations the same as Whites do is
usually wrong

undecided
usually right

-5-

In respect to Negroes, what do you believe others think is right or wrong:

36.

37.

38.

39.

40.

41.

42.

227

expect Whites to enjoy working with Negroes is

to resist arrest from White officials is

to be the victims of "police brutality" from Whites is

expect Whites to misuse trial-by-jury when they deal with Negroes is

To expect Whites to vote with Negroes for Negro candidates for public

expect Negroesto be given draft deferments equally with Whites is

To

1. usually wrong
2. undecided

3. usually right
To expect Negroes
1. usually right
2. undecided

3. usually wrong
To expect Negroes
1. usually right
2. undecided

3. usually wrong
To

1. usually right
2. undecided

3. usually wrong
office is

1. usually wrong
2. undecided

3 usually right
To

1. usually wrong
2. undecided

3. usually right

To

expect Negroes

usually wrong
undecided
usually right

to be as careful with their weapons as Whites are is

[2272

-7-

ABS-IV-WN-G

 

Directions: Section IV

 

This section contains statements about how you think you would act toward
Negroes. Choose the answer that indicates how you think you would act.

In respect to a Negro person would you yourself:

43.

44.

45.

46.

47.

48.

49.

 

I would trust Negroes with money

1. no
2. undecided
3. yes

I would want my family to be as closely knit as Negro families are

1. no
2. undecided
3. yes

I would want my intellectual ability to be the same as that of Negroes

1. no
2. undecided
3. yes

I would want to have the same desire Negroes do for a higher education

1. no
2. undecided
3. yes

I would help Negro neighbors

1. no
2. undecided
3. yes

I would want White neighborhoods to be as safe as I think Negro ones are

1. no
2. undecided
3. yes

I would obey job rules and regulations the same as Negroes

1. no
2. undecided
3. yes

2272

-3-

ABS-IV-WN-G

In respect to a Negro person would you yourself:

50.

51.

52.

53.

54.

55.

56.

I would enjoy working with Negroes

1. no
2. undecided
3. yes

I would resist arrest if arrested by Negroes

1. yes
2. undecided -
3. no 5}
l .
I would expect "police brutality" from Negroes .
l ..3-
1. yes
2. undecided
3. no

Would you use trial-by-jury equally when dealing with Negroes?

1. no
2. undecided
3. yes

Would you vote for a Negro candidate for public office?

1. no
2. undecided
3. yes

Would you want Whites to be given draft deferments as much as Negroes?

1. no
2. undecided
3. yes

WOuld you be as careful with weapons as you think Negroes are?

1. no
2. undecided
3. yes

2272

-9-

ABS-V-WN-G

 

Directions: Section V

 

This section concerns actual feelings that you yourself may have about Negroes.
You are asked to indicate how you feel about the following statements.

 

How do you actually feel toward Negroes:

57.

58.

59.

60.

61.

62.

63.

 

When Whites trust Negroes with money I feel

1. bad
2. indifferent
3. good

When White families are as closely knit as I think Negro families
are I feel

1. bad
2. indifferent
3. good

When Negroes' intellectual ability is the same as Whites I feel

1. bad
2. indifferent
3. happy

When Negroes desire a higher education as much as Whites do, I feel

1. bad
2. indifferent
3. good

When Whites help Negro neighbors I feel

1. bad
2. indifferent
3. good

When Whites are safe in Negro neighborhoods, I feel

1. bad
2. indifferent
3. good

When Negroes obey job rules and regulations the same as Whites, I feel

1. dissatisfied
2. indifferent
3. satisfied

.12272

-10-

ABs-V-WN-G

How do you actually feel toward Negroes:

64.

65.

66.

67.

68.

69.

70.

When Whites enjoy working with Negroes I feel

1. bad
2. indifferent
3. good

When Negroes resist arrest the same as Whites, I feel
1. dissatisfied

2. indifferent

3. satisfied

When Negroes use "police brutality" the same as Whites do, I feel

1. bad
2. indifferent
3. good

When Whites misuse trial-by-jury in relation to Negroes, I feel

1. happy
2. indifferent
3. angry

When Whites vote for Negro candidates for public office, I feel

1. bad
2. indifferent
3. good

When Negroes are given draft deferments as much as Whites, I feel

1. dissatisfied
2. indifferent
3. satisfied

When Negroes are as careful with their weapons as Whites are, I feel
1. dissatisfied

2. indifferent
3. satisfied

2712

-11-

ABS-VI-WN-G

 

Directions: Section VI

This section concerns actual experiences you have had with Negroes. Try to
answer the following questions from the knowledge of your own experience. If
you have had pg_experience or contact with Negroes,omit questions 71:84 and
begin again at question number 85. If you have had any experience or contact
with Negroes answer all of the following questions. '__

 

 

 

Experiences or contacts with Negroes:

 

71. I have trusted Negroes with money

1. no
2. uncertain
3. yes

72. I have seen that Negro families are as closely knit as White ones

1. no
2. uncertain
3. yes

73. The intellectual ability of Negroes is equal to mine

1. no
2. uncertain
3. yes

74. The Negroes I know wanted a higher education as much as I did

1. no
2. uncertain
3. yes

75. I have helped a Negro neighbor

1. no
2. uncertain
3. yes

76. I have felt safe when in Negro neighborhoods

1. no
2. uncertain
3. yes

77. I have seen that Whites obey job rules and regulations when working
with Negroes

1. no
2. uncertain
3. yes

2272

-12-

ABS-VI-WN-G

 

Experiences or contacts with Negroes:

 

78.

69.

80.

81.

82.

83.

84.

I have enjoyed working with Negroes

1. no
2. uncertain
3. yes

I have resisted arrest by Negroes

1. yes
2. uncertain
3. no

I have been the victim of "police brutality" from Negroes

1. yes
2. uncertain
3. no

I have seen Whites misuse trial-by-jury with Negroes

1. yes
2. uncertain
3. no

I have seen that Whites vote for Negro candidates for public office

1. no
2. uncertain
3. yes

I have seen that Negroes are given draft deferments as much as Whites

1. no
2. uncertain
3. yes

I have seen that Negroes are as careful with their weapons as Whites

1. no
2. uncertain
3. yes

-13- ABS-WN-D

ATTITUDE BEHAVIOR SCALE - ABS-WN-D
This part of the questionnaire deals with many things. For the purpose of

this study, the answers 2: all persons are important.

 

 

Part of the questionnaire has to do with personal information about you.
Since the questionnaire is completely anonymous or confidential, you may
answer all of the questions freely without any concern about being identified.

15 is_ipportant go the study so obtain your answer £o_every question.

Please read each question carefully and_g2 not omit any questions. Please

answer by circling the answer or marking the space on the IBM answer sheet.

85. Please indicate your sex.
1. Female
2. Male

86. Please indicate you age as follows:

1. Under 20
2. 21-30
3. 31-40
4. 41-50

5. Sl-over
87. What is your marital status?
1. Married
2. Single
3. Divorced
4. Widowed

5. Separated

.2272

-14- ABS-WN-D

88. What is your religion?

1.

2.

3.

4.

5.

I prefer not to answer
Catholic

Protestant

Jewish

Other

89. Please indicate level of education

4.

5.

A

First year university

Second year university

l-r - \- —A\ -.‘h- .4 ‘fi
I. u
I

Third year university
Fourth year university

Graduate student

90. Some people feel that in bringing up children, new ways and methods should
be tried whenever possible. Others feel that trying out new methods is
dangerous. What is your feeling about the following statement?

"New methods of raising children should be tried whenever possiblef'

l.

2.

3.

4.

Strongly disagree
Slightly disagree
Slightly agree

Strongly agree

91. Family planning on birth control has been discussed by many people. What
is your feeling about a married couple practicing birth control?

1.

2.

112272

It is always wrong
It is usually wrong
It is probably all right

It is always right

-15-

ABS-WN-D

92. The following questions have to do with kinds of eXper1ences you have had

with Negroes.

answer with the highest number.

1.

If more than one experience applies, please choose the

I have read or studied about Negroes through reading, movies, lecture

or observation.

A friend or relative is a Negro person.

I have personally worked with Negroes as teacher, counselor, volunteer,

child care, etc.

93. Considering all of the times you have talked, worked or in some other way
had personal contact with Negroes about how much has it been altogether?

1.

2.

3.

4.

5.

Only a few casual contacts.
Between one and three months.
Between three and six months.
Between six months and one year.

More than one year of contact.

r=s

[‘I' -. ‘ .‘

94. When you have been in contact with Negroes, how easy for you, in general,
would you say it would have been to have avoided being with them?

1.

2.

I have had no contact.

I could generally have avoided these
cost or difficulty.

I could generally have avoided these
considerable difficulty.

I could generally have avoided these
inconvenience.

I could generally have avoided these
difficulty or inconvenience.

personal

personal

personal

personal

contacts only at great

contacts only with

contacts but with some

contacts without any

95. If you have ever worked with Negroes for personal gain (for example, for
money or some other gain) what opportunities did you have (or do you have)
to work at something else instead; that is, something else that was (is)
acceptable to you as a job?

1.

2.

112272

No such experience.

No other job available.

Other jobs available not s5 all acceptable to me.

Other jobs available were not guite acceptable to me.

Other jobs available were fully acceptable to me.

-16- ABS-WN-D

96. How have you generally felt about your experiences with Negroes?

l.

2.

3.

4.

5.

No experience.

I definitely dislike it.

I did not like it very much.
I like it somewhat.

I definitely enjoyed it.

97. Which of the following do you think would have the greatest effect of
reducing racial prejudice? Circle only one or mark one on the IBM
answer sheet.

1. Integration of schools.

2. Publicity campaigns to promote integration.

3. Fair employment legislation.

4. Open housing legislation.

5. Direct, personal contact between members of various racial groups.
98. How would you rate your own racial attitudes as compared to the average

person?

1. Very much more prejudiced.

2. Somewhat more prejudiced.

3. About the same.

4. Somewhat less prejudiced.

5. Very much less prejudiced.

99. To which racial group do you belong?

l.

2.

112272

Prefer not to answer.

White

Black

Oriental

Other

6 C1 u-Ml'1

-l7- ABS-WN-D
100. Where were you mainly reared or "brought up” in your youth (that is, up
to age 21)?
1. country
2. country town
3. city suburb

4. city

 

112272

-18- ABS-VJ-D

LIFE SITUATIONS

This section of the booklet deals with how people feel about several aspects of
life or life situations. Please indicate how you feel about each situation by
circling the answer you choose or marking on the IBM answer sheet.

101.

102.

103.

104.

105.

106.

112272

It should be possible to eliminate war once and for all

DwNI-I
0..

strongly
disagree
agree

strongly

disagree

agree

Success depends to a large part on luck and fate 5'1

waI-I
0..

strongly
agree

disagree
strongly

Someday most

buNI-I

By

strongly
disagree
agree

strongly

improving

in the world

waI—I

strongly
disagree
agree

strongly

agree

disagree
of the mysteries of the world will be revealed by science

disagree

agree

industrial and agricultural methods, poverty can be eliminated

disagree

agree

With increased medical knowledge it should be possible to lengthen the
average life span to 100 years or more

waI—I

strongly disagree

disagree
agree
strongly

agree

Someday the deserts will be converted into good farming land by the
application of engineering and science

Il-‘wNI—I
.0.

strongly disagree

disagree
agree
strongly

agree

112272

107.

108.

109.

Education can only help people develop their natural abilities; it

-19- ABS-WN-D

cannot change people in any fundamental way.

waH
0.

strongly
agree

disagree
strongly

agree

disagree

With hard work anyone can succeed

«bump—I
0..

strongly
disagree
agree

strongly

Almost every

J-‘UONH
0

strongly
disagree
agree

strongly

disagree

agree
present human problem will be solved in the future

disagree

agree

-_. Man-—

"I”

APPENDIX B-3

ATTITUDE BEHAVIOR SCALE BW-G

334

ATTITUDE BEHAVIOR SCAIE-BW-G

DIRECTIONS

 

This booklet contains statements of how people behave in certain situations
or feel about certain things. You, yourself, or other Black persons often
behave in the same way toward Whites. You also have some general ideas
about yourself, about other Black persons like you and about Whites.
Sometimes you feel or behave the same way toward everyone and sometimes you
feel or behave differently toward Whites.

 

This questionnaire has statements about ideas and about behavior. Each
statement of this questionnaire is different from every other section,
although some of the statements in each section are similar. Your answers
in one section, therefore, may be the same as answers in another section,
or your answers may differ from section to section. Here is a sample
statement:

Sample‘l

Other Blacks believe the following things about Whites as compared to Blacks:

 

l. Chance of Whites being taller

9 less chance than Blacks
. about the same
3. more chance than Blacks

If other Blacks believe that Whites have less chance than Blacks to be
taller, you should circle the number 1 as shown above or if you are using
an IBM answer sheet make a heavy dark line on the answer sheet between the
two lines after the number as follows:

 

 

1,1- 2:: 3:2: 4:: 5::

******************* DO NOT PUT YOUR NAME ON THE BOOKLET *******************

by: (:kichard J. Hamersma
John E. Jordan
College of Education
Michigan State University

112272

ABS-I-BW-G
Directions: Section I
This section contains statements about ideas which other Blacks have about

Whites. Circle 2; fill in the answer sheet number that indicates how
other Blacks compare Whites with themselves. Please answer all questions.

 

Other Blacks believe the following things about Whites as compared to Blacks:
1. Whites can be trusted with money

1. less than Blacks

2. about the same as Blacks

3. more than Blacks
2. White families are closely knit

1. less often than Black ones

2. about as often as Black ones
3. more often than Black ones

3. Whites' intellectual ability is

1. less than Blacks'
2. about the same as Blacks'
3. more than Blacks'

4. Whites desire a higher education
1. less often than Blacks

2. about as often as Blacks
3. more often than Blacks

U1
0

Whites help their neighbors

1. less than Blacks do

2. about the same as Blacks do

3. more than Blacks do

6. White neighborhoods are safe
1. less often than Black ones
2. about as often as Black ones
3. more often than Black ones

7. Whites obey job rules and regulations

1. less than Blacks

2. about the same as Blacks

3. more than Blacks

8. Blacks enjoy working with Whites

1. less than Whites do with Blacks
2. about the same as Whites
3. more than Whites do with Blacks

117777

-2-
ABS-I-BW-G

()ther Blacks believe the following things about Whites as compared to Blacks:

 

9. Whites resist arrest
1. more than Blacks
2. about the same as Blacks
3. less than Blacks

10. Whites are victims of "police brutality"
1. more than Blacks
2. about the same as Blacks
3. less than Blacks

11. Whites misuse trial-by-jury
1. more often than Blacks
2. about as often as Blacks
3. less often than Blacks

12. Blacks vote for White candidates for public office
1. less than Whites do for Blacks
2. about the same as Whites do for Blacks
3. more than Whites do for Blacks

13. Whites desire draft deferments
l. more often than Blacks
2. about as often as Blacks
3. less often than Blacks

14. Whites are careful with their weapons
1. less often than Blacks

2. about as often as Blacks
3. more often than Blacks

112272

.r- -—r-—-q

-3-

ABS-II-BW-G

 

Directions: Section II

This section contains statements about things which most Blacks generally
believe others would experience when interacting with Whites. Please choose
the answer that indicates what you think most others believe about Whites.

 

 

Most Blacks generally believe the following about interacting with Whites:

 

15. Blacks believe they can trust Whites with money

1. disagree
2. uncertain
3. agree HI‘

16. Blacks believe that White families are as closely knit as their own i
1. disagree e..
2. uncertain
3. agree

17. Blacks believe the intellectual ability of Whites is equal to theirs

1. disagree
2. uncertain
3. agree

18. Blacks believe they want to do their study or school work with Whites

1. disagree
2. uncertain
3. agree

19. Whites like to help their neighbors
1. less than Blacks do
2. about the same
3. more than Blacks do
20. Blacks believe that White neighborhoods are safe for Blacks
1. disagree
2. uncertain
3. agree
21. Blacks believe Whites obey job rules and regulations as much as they do
1. disagree

2. uncertain
3. agree

112272

—q—

ABS-II-BW-G

 

Most Blacks generally believe the following about interacting with Whites:

 

22.

23.

24.

25.

26.

27.

28.

Blacks believe they enjoy working with Whites

1. disagree
2. uncertain
3. agree

Whites believe in resisting arrest from Black policemen
1. agree
2. uncertain

3. disagree

Blacks believe that Whites are the victims of "police brutality"
from Blacks

1. agree
2. uncertain
3. disagree

Blacks usually use trial-by-jury fairly when they deal with Whites

1. usually not believed in
2. undecided
3 usually believed in

For Blacks to vote with Whites for White candidates for public office
1. usually not approved

2. undecided

3 usually approved

For Blacks to be given draft deferments more than Whites

1. usually approved

2. undecided

3 usually not approved

For Whites to be as careful with their weapons as Blacks are

1. usually not approved
2. undecided
3. usually approved

112272

IH rut-Lions: Section III

 

-J-

ABS-III-BW-C

 

This section contains statements of the right or wrong way of behaving or acting

toward Whites.

You are asked to indicate what you yourself belie!g_others

 

think should be done with respect to Whites.

In reapect to Whites, what do 222 believe othoza think is right

expect White families to be as closely knit as Black ones is

expect Whites' intellectual ability to be the same as Blacks is

desire a higher education as much as Blacks do is

help White neighbors is

believe that White neighborhoods are safe for them is

obey job rules and regulations the same as Blacks do is

or wrong:
29. To trust Whites with money is
1 usually wrong
2. undecided
3 usually right
30. To
1. usually wrong
2. undecided
3. usually right
31. To
1. usually wrong
2. undecided
3. usually right
32. To expect Whites to
1. usually wrong
2. undecided
3. usually right
33. To expect Blacks to
1. usually wrong
2. undecided
3. usually right
34. To expect Blacks to
1. usually wrong
2. undecided
3. usually right
35. To expect Whites to
1. usually wrong
2. undecided
3. usually right

112272

-b-

ABS-III-BW-G

 

In respect to Whites, what do you believe others think is right or wrong:

36.

37.

38.

39.

40.

41.

42.

To expect Blacks to

1. usually wrong
undecided
3. usually right

N

To expect Whites to

usually right
undecided
usually wrong

UNH

To expect Whites to

usually right
undecided
usually wrong

“NH

To expect Blacks to

1. usually right
2. undecided
3. usually wrong

To expect Blacks to
office is

1. usually wrong
2. undecided
3. usually right

To expect Whites to
1. usually wrong
2. undecided

3. usually right
To expect Whites to
1. usually wrong

2. undecided
3. usually right

112272

enjoy working with Whites is

resist arrest from Black officials is

be the victims of "police brutality" from Blacks is

misuse trial-by-jury when they deal with Whites is

vote with Whites for White candidates for public

be given draft deferments equally with Blacks is

be as careful with their weapons as Blacks are is

ABS-IV-BW-G

 

Directions: Section IV

This section contains statements about how you think you would act toward
Whites. Choose the answer that indicates how you think you would act.

In reapect to a White person would you yourself:

43.

44.

45.

46.

47.

48.

49.

I would trust Whites with money

1. no

2. undecided

3. yes

I would want my family to be as closely knit as White families are

1. no

2 undecided

3. yes

I would want my intellectual ability to be the same as that of Whites
1. no

2 undecided

3. yes

I would want to have the same desire Whites do for a higher education
1. no

2 undecided

3. yes

I would help White neighbors

1. no

2 undecided

3. yes

I would want Black neighborhoods to be as safe as I think White ones are
1. no

2 undecided

3. yes

I would obey job rules and regulations the same as Whites

1. no

2. undecided

3. yes

112272

-3-

ABS-IV-BW-G

 

In respect to a White person would you yourself:

50.

51.

52.

S3.

54.

SS.

56.

112272

I would enjoy working with Whites
1. no

2. undecided

3. yes

I would resist arrest if arrested by Whites

1. yes
2. undecided
3. no

I would expect "police brutality" from Whites

1. yes i
2. undecided {
3. no

Would you use trial-by-jury equally when dealing with Whites?

1. no
2. undecided
3. yes

Would you vote for a White candidate for public office?

1. no
2. undecided
3. yes

Would you want Blacks to be given draft deferments as much as Whites?

1. no
2. undecided
3. yes

Would you be as careful with weapons as you think Whites are?

1. no
2. undecided
3. yes

-’-

ABS-V-BW-C

 

Directions: Section V

 

This section concerns actual feelings that you yourself may have about Whites.
You are asked to indicate how you feel about the following statements.

 

 

How do you actually feel toward Whites:

 

57. When Blacks trust Whites with money I feel

1. bad
2. indifferent
3. good

58. When Black families are as closely knit as I think White families
are I feel

1. bad
2. indifferent
3. good

59. When Whites' intellectual ability is the same as Blacks I feel

1. bad
2. indifferent
3. happy

60. When Whites desire a higher education as much as Blacks do, I feel

1. bad
2. indifferent
3. good

61. When Blacks help White neighbors I feel

1. bad
2. indifferent
3. good

62. When Blacks are safe in White neighborhoods, I feel

1. bad
2. indifferent
3. good

63. When Whites obey job rules and regulations the same as Blacks, I feel
1. dissatisfied

2. indifferent
3. satisfied

112272

-10-

ABS-V-BW-G

 

How do you actually feel toward Whites:

64.

65.

66.

67.

68.

69.

70.

 

When Blacks enjoy working with Whites I feel

1. bad
2. indifferent
3. good

When Whites resist arrest the same as Blacks, I feel
1. dissatisfied

2. indifferent

3. satisfied

When Whites use "police brutality" the same as Blacks do, I feel

1. bad
2. indifferent
3. good

When Blacks misuse trial-by-jury in relation to Whites, I feel

1. happy
2. indifferent
3. angry

When Blacks vote for White candidates for public office, I feel

1. bad
2. indifferent
3. good

When Whites are given draft deferments as much as Blacks, I feel

1. dissatisfied
2. indifferent
3. satisfied

When Whites are as careful with their weapons as Blacks are, I feel
1. dissatisfied

2. indifferent
3. satisfied

112272

AISS-VI- ISW-(t

 

Directions: Section VI

 

This section concerns actual Experiences you have had with Whites. Try to
answer the following questions from the knowledge of your oyp_experience. If
you have had 33 experience or contact with Whites, omit questions 71-84 and
begin again at question number 85.. If you have had any experience 25 contact
with Whites answer all of the following questions.

 

 

 

 

Experiences or contacts with Whites:

 

71.

72.

73.

74.

75.

76.

77.

I have trusted Whites with money
1. no

2. uncertain

3

yes

I have seen that White families are as closely knit as Black ones

1. no
2. uncertain
3. yes

The intellectual ability of Whites is equal to mine

1 no
2. uncertain
3 yes

The Whites I know wanted a higher education as much as I did

1 no
2. uncertain
3 yes

I have helped a White neighbor
1. no

2. uncertain

3

yes

I have felt safe when in White neighborhoods

1. no
2. uncertain
3. yes

I have seen that Blacks obey job rules and regulations when working
with Whites

1. no
2. uncertain
3. yes

112272

-12-

ABS-VI-BW-G

 

Experience or contacts with Whites:

 

78.

79.

80.

81.

82.

83.

84.

I have enjoyed working with Whites

1. no
2. uncertain
3. yes

I have resisted arrest by Whites

1. yes
2. uncertain
3. no

I have been the victim of "police brutality" from Whites

1. yes
2. uncertain
3. no

I have seen Blacks misuse trial-by-jury with Whites

1 yes
2. uncertain
3 no

I have seen that Blacks vote for White candidates for public office
1. no

2. uncertain

3

yes

I have seen that Whites are given draft deferments as much as Blacks

1. no
2. uncertain
3. yes

I have seen that Whites are as careful with their weapons as Blacks

1. no
2. uncertain
3. yes

112272

112272

-13-

ATTITUDE BEHAVIOR SCALE - ABS-BW-D

This part of the questionnaire deals with many things. For the purpose of this

study, the answers pf all persons are important.

 

 

Part of the questionnaire has to dO‘With personal information about you. Since
the questionnaire is completely anonymous or confidential, you may answer all of
the questions freely without any concern about being identified. lE.i§ important

£2_the study £9 obtain your answer £2 every question.

 

 

Please read each question carefully and d2 not omit any questions. Please answer

by circling the answer or marking the peace on the IBM answer sheet.

85. Please indicate your sex.
1. Female
2. Male
86. Please indicate your age as follows:

1. Under 20

2. 21-30
3. 31-40
4. 41-50

5. 51-over
87. What is your marital status?
1. Married
2. Single
3. Divorced
4. Widowed

5. Separated

112272

88.

89.

90.

91.

~14- ABS-BW-D

What is your religion?

1. I prefer not to answer

2. Catholic

3. Protestant

4. Jewish

5. Other

Please indicate level of education

1. First year university

2. Second year university

3. Third year university

4. Fourth year university

5. Graduate student

Some people feel that in bringing up children, new ways and methods should
be tried whenever possible. Others feel that trying out new methods is
dangerous. What is your feeling about the following statement?

"New methods of raising children should be tried whenever possible."

1. Strongly disagree

2. Slightly disagree

3. Slightly agree

4. Strongly agree

Family planning on birth control has been discussed by many people. What
is your feeling about a married couple practicing birth control?

1. It is always wrong
2. It is usually wrong
3. It is probably all right

4. It is always right

112272

92.

93.

94.

95.

-15- ABS-BW-D

The following questions have to do with kinds of experiences you have had
with Whites. If more than one experience applies, please choose the answer
with the highest number.

 

 

1. I have read or studied about Whites through reading, movies, lecture
or observation.

2. A friend or relative is a White person.

3. I have personally worked with Whites as a teacher, counselor, volunteer,
child care, etc.

Considering all of the times you have talked, worked or in some other way
had personal contact with Whites, about how much has it been altogether?

1. Only a few casual contacts.

2. Between one and three months.

3. Between three and six months.

4. Between six months and one year.
5. More than one year of contact.

When you have been in contact with Whites, how easy for you, in general,
would you say it would have been to have avoided being with them?

1. I have had no contact.

2. I could generally have avoided these personal contacts only at great
cost of difficulty.

3. I could generally have avoided these personal contacts only with
considerable difficulty.

 

4. I could generally have avoided these personal contacts but with some
inconvenience.

5. I could generally have avoided these personal contacts without any
difficulty or inconvenience.

 

If you have ever worked with Whites for personal gain (for example, for
money or some other gain) what opportunities did you have (or do you have)
to work at something else instead; that is, something else that was (is)
acceptable to you as a job?

1. No such experience.
2. No other job available.

3. Other jobs available not at all acceptable to me.

4. Other jobs available were not quite acceptable to me.

5. Other jobs available were fully acceptable to me.

112272

96.

97.

98.

99.

-16- ABS-BW-D

How have you generally felt about your experiences with Whites?

1. No experience.

2. I definitely dislike it.

3. I did not like it very much.

4. I like it somewhat.

5. I definitely enjoyed it.

Which of the following do you think would have the greatest effect of
reducing racial prejudice? Circle only on or mark only one on the IBM
answer sheet.

1. Integration of schools.

2. Publicity campaigns to promote integration.

3. Fair employment legislation.

4. Open housing legislation.

5. Direct, personal contact between members of various racial groups.

How would you rate your own racial attitudes as compared to the average
person?

1. Very much more prejudiced.

2. Somewhat more prejudiced.

3. About the same.

4. Somewhat less prejudiced.

5. Very much less prejudiced.

To which racial group to you belong?
1. Prefer not to answer.

2. White

3. Black

4. Oriental

5. Other

-17- ABS-BW-D

100. Where were you mainly reared or "brought up" in your youth (that is up to
age 21)?

1. Country
2. Country town
3. City suburb

4. City

112272

rm—t—.

- l8- ABS-8W4)

LIFE SITUATIONS

This section of the booklet deals with how people feel about several aspects of
life or life situations. Please indicate how you feel about each situation by
circling the answer you choose or marking on the IBM answer sheet.

101.

102.

103.

104.

105.

106.

112272

It should be possible to eliminate war once and for all

4‘me
0.0

0

strongly disagree

disagree
agree
strongly

agree

Success depends to a large part on luck and fate

Dri—i

strongly
agree

disagree
strongly

Someday most

wat-I

BY

strongly
disagree
agree

strongly

improving

the world

J-‘LONr-t

strongly
disagree
agree

strongly

agree

disagree
of the mysteries of the world will be revealed by science

disagree

agree

industrial and agricultural methods, poverty can be elimdnated in

disagree

agree

With increased medical knowledge, it should be possible to lengthen the
average life span to 100 years or more

bump-t

strongly
disagree
agree

strongly

disagree

agree

Someday the deserts will be converted into good farming land by the application
of engineering and science

war—I
0

strongly
disagree
agree

strongly

disagree

agree

-19- ABS-BW-U

107. Education can only help people develop their natural abilities; it
cannot change people in any fundamental way

1. strongly agree

2. agree

3. disagree

4. strongly disagree

108. With hard work anyone can succeed

strongly disagree
disagree

agree

strongly agree

war-I
O.

109. Almost every present human problem will be solved in the future

1. strongly disagree
2. disagree

3. agree

4

strongly agree

112272

APPENDIX C

SEMANTIC DIFFERENTIAL

355

 

j, u r

SEMANTIC DIFFERENTIAL

DIRECTIONS

The purpose of this study is to measure the meanings of certain

 

things to various peOple by having them judge them against a
series of descriptive scales. In taking this scale, please make
your judgments on the basis of what these things mean to you. On
each page, you will find a different concept to be judged and
beneath it a set of scales. You are to rate the concept on each

of these scales in order.

Here is how you are to use these scales: If you feel that the
concept at the top of the page is very closely related to one end
of the scale, you should place your checkmark as follows:

fair X : : : : : : unfair

 

or

fair : : : : : : X unfair

 

If you feel that the concept is quite closely related to one or the
other end of the scale (but not extremely), you should place your
check-mark as follows:
strong : X : : : : : 'weak
or

strong : : : : : X : weak

 

356

357

If the concept seems only slightly related to one side as opposed
to the other side (but is not really neutral), then you should check

as follows:

 

active : : X : : : : passive
or
active : : : : X : : passive

 

The direction toward which you check, of course, depends upon which
of the two ends of the scale seem most characteristic of the think
you're judging. If you consider the concept to be neutral on the
scale, both sides of the scale eggally associated with the concept,
or if the scale is completely irrelevant, unrelated to the concept,
then you should place your check-mark in the middle space:
safe : : : X : : : dangerous
IMPORTANT: (1) Place your checkrmarks in the middle of spaces,
not on the boundaries:
This Not this
: : : X : X °
(2) Be sure you check every scale for every concept--
do not omit any.
(3) Never put more than one check-mark on a single
scale.

Sometimes you may feel as though you've had the same item before on
the test. This will not be the case, so do not look back and forth
through the items. Do not try to remember how you checked similar
items earlier in the test. Make each item a separate and independent
judgment. Work at a fairly high speed through this test. Do not
worry or puzzle over individual items. It is your first impressions,
the immediate "feelings" about the items, that we want. On the other

hand, please do not be careless, because we want your true impressions.

kind
thick
clean
cold
had

358

FRIEND

cruel
thin
dirty
hot

 

 

 

strong
honest
active

good

weak
dishonest
passive

 

unpleasant
small

 

pleasant

 

large

 

worthless
slow

0. O. O. O. O. O. I. 0. .0 CI. 0.

 

valuable

 

fast

 

fair
soft

 

unfair
hard

 

cowardly

 

sharp
friendly

 

heavy

brave

 

dull

 

 

 

 

unsuccessful

 

unfriendly

light

 

 

violent

O. O. O. C. I. O. O. I. O.

 

 

O. O. O. O. O. O. O. O. O. O. O. O. 00 I. O. O. O. O. O. O.
O. O. to O. O. n O. O. O. O. O. O. O. .0 O. .0 O. O. O. O.
O. O. O. O. O. O. O. O. O. O. I. .0 O. 0. O. O. O. O. O. O.

O. O. O. O. O. O. O. O. .0 DO 0. O. O. 0. co 0. .0 O. O. O.
O. O. O. O. O. O. O. O. O. O. O. 0. O. 0. on O. O. .0 O. 00

successful

ENEMY

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

kind a : z z x : cruel
thick : y__: x z : : thin
clean : _p: x 3 z 2 dirty
cold : x z z 3 : hot
bad x x x x 8 2 good
strong : z x z 2 3 weak
honest : z : x x :_> dishonest
active x z z x z : passive
unpleasant : s : z : : pleasant
small : z z z x 2 large
worthless : x : : z : valuable
slow : : x : : : fast
fair 2 z x z z : unfair
soft : z z 8 x : hard
cowardly : x z 2 x : brave
sharp : 2 s z x : dull
friendly : x z x z : unfriendly
heavy : : a z z : -light
unsuccessful : z z z : : successful
violent : z z : x 3 moderate

 

360

BLACK PERSON

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

kind 2 2 2 2 2 2 cruel
thick 2 2 2 2 2 2 thin
clean 2 2 2 2 2 2 dirty
cold 2 2 2 2 2 2 hot
bad 2 2 2 2 2 2 good
strong 2 2 2 2 2 : weak
honest 2 2 2 2 2 2 dishonest
active 2 2 2 2 2 2 passive
unpleasant 2 2 2 2 2 2 pleasant
small 2 2 2 2 2 2 large
worthless 2 2 2 2 2 2 valuable
slow 2 2 2 2 2 2 fast
fair 2 2 2 2 2 2 unfair
soft 2 2 2 2 2 2 hard
cowardly 2 2 2 2 2 2 brave
sharp 2 2 2 2 2 2 dull
friendly 2 2 2 2 2 2 unfriendly
heavy 2 2 2 2 2 2 light
unsuccessful 2 2 2 2 2 2 successful
violent 2 2 2 2 2 2 moderate

 

 

kind

361

WHITE PERSON

cruel

 

thick

 

clean

cold

bad
strong

honest

active
unpleasant
small

I. O. O. O. O. O. O. O. O. .0

 

thin

dirty

 

hot

good

weak
dishonest
passive
Apleasant
large

 

worthless
slow
fair
soft

valuable
fast
unfair
hard

 

cowardly
sharp
friendly

 

heavy

brave
dull

unfriendly

light

 

unsuccessful
violent

00000000000000.0000.

”C.“”OO”OOOOHOOOOOOCOOOOO~OOOO”OO

N”0.000000000000COOOOO“OOOQOOOOOOOOOO
”MOOOOOOOO”OOOOOOOOOOOOOCOOOOOOOOOOOO

”““”-OOOOOOOOCO“-OOOOOOOO~”

8
8
8
8
8
8
8
8
8
8
3
8
8
8
8
8

successful
moderate

kind

362

NEGRO PERSON

cruel

 

thick

thin

 

clean
cold
bad

9 O O. 0.

dirty
hot
good

 

strong

weak

 

honest

dishonest

 

active

0. O. O. O. O.

ppassive

 

unpleasant

pleasant

 

small

large

 

worthless
slow

valuable
fast

 

fair

 

soft

unfair

 

hard

 

cowardly

brave

 

sharp

o-oeuoouoouuuuoouoooooou

dull

 

friendly
heavy

O. O. O. O. O. M O. O. 0. O. O. O. O. O. O. O. O.

unfriendly
light

 

unsuccessful
violent

so 0. O. O. O. O. O. O. O. O. G. O. O. 0. O. O. O. I. O. D.

O. O. O. O. O. O. O. O. O. O. I. O. O. O. O. O. O. O. O. u

0. O. O. O. O. O. O. O. O. 0. O. ‘.

O. O. O. O. O. O. O. O. O. .0 O. O. O. O. O. O. 0. O. O. O.

successful

moderate