M

H
| I:

r}!
I

Hit
.HI

W

w

I

W
‘
I

“HI

|

 

EM?ACF Q? CQN?ACF CH
MCEchEYHNEC ATHWQE

Thesis 509 {“m Duqmo a‘: M. A.
{\‘ihCziiGz'J-LR STATE UN?! 13:73:21“?

Kathryn Laws
3087

   
    

.' A j {t ‘ , 13 J.
THES‘S M f." ’~}(& I‘\ \‘\‘ [if ‘ ,i" f,

  

LIBRARY

Michigan Static
University

  

 

—-"/'¢

 

ABSTRACT

IMPACT OF CONTACT ON RACIO-ETHNIC ATTITUDES

by Kathryn Lass

This paper attempts to deal with relationships that
may exist between the variable of previous prolonged person-
al contact with members of other races and attitudes of
racio-ethnic liberalism toward those races. Specifically,
the general hypothesis tested is: "Those who have had previ-
ous prolonged personal contact with members of other races
exhibit a higher degree of racio-ethnic liberalism and
social acceptance than do those who have had no previous pro—
longed personal contact with members of other races."

Using data from 265 questionnaires administered to
college students, the hypothesis was supported (at a .05
level or below, using chi—square analysis) for certain social
and demographic categories of the controls applied. These
categories are females, those from small towns, those with
aihighfrequency of church attendance, those whose fathers
are in the middle status occupational group, those with low
grade points, those of a moderate political orientation, and

those above freshman status.

Kathryn Lass

An attempt was also made to identify some correlates
of racio-ethnic prejudice. That is, what groups are more
prejudiced, regardless of previous contact? Here, the groups
that exhibited the greatest degree of prejudice most con-
sistently are the same groups, with the exception of college
status and political orientation, as those for whom contact
was a statistically significant factor in their racio-ethnic
attitudes. This suggests that when people previously hold
strong prejudices, personal contact with the social objects
of their prejudices results in a significant reduction of
same. For those who previously hold weaker prejudices,
contact would not be as significant a factor in their attitude
formation.

A peripheral concern was to discover the differential
impact of type of contact on attitudes. Generally speaking,
it was found that those who had experienced contact with
members of other races on a peer basis were the most racio-
ethnically liberal, those who had experienced contact on a
dominant-subordinate level were the next most liberal, and
those who had experienced no contact at all were the least

racio-ethnically liberal.

IMPACT OF CONTACT ON RACIO-ETHNIC ATTITUDES

BY

Kathryn Lass

A THESIS

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

MASTER OF ARTS

College of Social Science

1967

ACKNOWLEDGEMENTS

Although I, of course, assume full responsibility for
all errors of omission and commission which may be contained
in this paper, I would like to gratefully acknowledge the
aid of the following people in the inception, progress, and
completion of this study.

To my thesis advisor, Dr. Frederick B. Waisanen, I owe
a debt of gratitude for his considerable guidance and intel-
lectual support.

To my colleagues in the graduate program of sociology
at Michigan State University, I wish to give sincere thanks
for their encouragement, both personal and intellectual.
Considerations of space preclude my listing each and every
one individually, but I hope that they will accept my sincere
expression of gratitude. Just the same, however, I would
like to especially thank Arnold Holden for giving of his

help in times of need.

ii

II.

III.

IV.

V.

VI.

TABLE OF CONTENTS

GENERAL PROBLEM AND BACKGROUND. . . . . . . . .
SPECIFIC PROBLEMS . . . . . . . . . . . . . . .

A. Contact. . . . . . . . . . . . . . . . . .
B. Configuration of Racio-ethnic Prejudice or

Ethnocentrism. . . . . . . . . . . . . . .
C. Type of Contact. . . . . . . . . . . . . .

METHODOLOGY . O O O O O O O O O O O O O O O O O

A. Operationalization of Variables. . . . . .
1. Contact--Independent Variable . . . .

2. Racio-ethnic Liberalism--Dependent
Variable. . . . . . . . . . . . . . .
5. Control Variables . . . . . . . . . .
B. Research Population and Data Collection. .

FINDINGS
A. Association Between Previous Contact and
RaCiO-ethnic Liberalism. o o o o o o o o o
B. Identification of Some Correlates of Racio-
ethnic Prejudice . . . . . . . . . . . . .
C. Type of Contact. . . . . . . . . . . . . .
SUMMARY AND DISCUSSION. . . . . . . . . . . . .

CONCLUSION. . . . . . . . . . . . . . . . . . .

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . .

iii

Page

(Ox)

10

10
10

11

14
17

18

25
55

56
46

48

LIST OF TABLES

TABLE

I. Chi-square Significance of Predicted Relation-
ship Between Contact and Liberalism (Numbers
1’ 2 and 5) O O O O O O O O O O O O O O O O O C

II. Chi-square Significance of Predicted Relation-
ship Between Contact and Liberalism (Numbers
4 and 5) O O O O C O O C O C O O C O O O O O 0

III. Chi-square Significance of Predicted Relation-

ship Between Contact and Liberalism (Numbers
6 and 7) . . . . . . . . . . . . . . . . . . .

iv

Page

20

22

24

I. GENERAL PROBLEM AND BACKGROUND

The greater the part that legislation has come to play
in implementing the processes of racial desegregation and
integration, the stronger have become the arguments that
"you can't legislate morality" and "you can't change people's
feelings through laws." To a degree, this has been demon-
strated by the emergence of strong reactionary elements,
politically referred to as the "white backlash," following,
for example, the 1954 Supreme Court decision ordering racial
integration of schools and the 1965 Civil Rights Act.

One can look at the situation in terms of just how, and
to what extent, people adjust themselves to rapid and, to
them, often disruptive, changes in the world around them.

The concern of attitude change is significant to the broad
area of social change--significant because of the pragmatic
implications that research in this area might have. Studies
in attitude change and attitude formation, especially concern-
ing inter-racial and inter-ethnic relations, have been abun-
dant.

Often, this type of research is intended to provide
guidelines for social policy. That is, it is hoped that there
is some element or variable which, when manipulated in a cer-

tain way, will effect a reduction of prejudice on the part of

the persons upon whom this variable is operative. Inter—
racial and inter-ethnic contact is one of these variables.
For example, Deutsch and Collins did a comparative study
of two relatively segregated housing projects (which provided
few opportunities for interracial contact) and two completely

1 In the integrated projects they

integrated housing projects.
found the housewives significantly less prejudiced; there was
no evidence that white housewives in the integrated projects
had been less prejudiced before coming to live in the housing
area than those in the more segregated projects.2 It is also
significant to note that, among the housewives in the segre-
gated project, their attitudes were more often less favorable
to interracial housing than they were before they moved in.
Allport concludes from his studies of prejudice that,
"The trend of evidence favors the conclusion that knowledge
about and acquaintance with members of minority groups make
for tolerant and friendly attitudes."3
In one study of ethnic and racial prejudice in six

American communities, it is concluded that, "In all the sur- .

veys in all cOmmunities and for all groups, majority and

 

1Morton Deutsch and Mary Collins, Interracial Housing
(Minneapolis: University of Minnesota Press, 1951).

2Theodore M. Newcomb, Ralph H. Turner, and Philip E.
Converse, Social Psychology (New York: Holt, Rinehart and
Winston, Inc., 1965), p. 87.

3Gordon W. Allport, The Nature of Prejudice (New York:
Doubleday and Co., Inc., 1958), p. 254.

minorities, the greater the frequency of interaction, the
lower the prevalence of ethnic prejudice."l

When white sailors in the Merchant Marine were studied,
a strong inverse relationship was found between the number
of times a sailor has shipped with Negroes and his anti-
Negro prejudices.2

These and numerous other studies strongly suggest that
contact highly correlates with, if it does not actually pro—
duce, favorable attitudes toward members of various ethnic

and racial groups.

 

1Robin M. Williams, Jr., Strangers Next Door (Englewood
Cliffs, New Jersey: Prentice-Hall, Inc., 1964), p. 168.

EIra N. Brophy, "The Luxury of Anti-Negro Prejudice,"
Public Opinion Quarterly, IX (Winter, 1945-46), pp. 456-466.

II. SPECIFIC PROBLEMS
A. Contact

In his classic work, The Human Group, Homans hypother
sized that, other things being equal, "If the frequency of
interaction between two or more persons increases, the degree
of their liking for one another will increase and vice-
versa."1

The most crucial "other thing," Homans feels, is author-
ity--in other words, a difference in the status positions of
the parties involved in the interaction. For, "The relation-
ship between inferior and superior in a hierarchy is inevit-
ably tension producing, and the conflict between different
people is always subject to interpretation in group terms."2
So perhaps it is the quality of the interaction rather than
the quantity of it that is crucial; it could be that inter-
action and unprejudicial attitudes (or sentiments of liking)
are positively associated only if a status differentiation is
not a factor inherent in the structure of the relationship.

This concern is brought out primarily because the nature

of the data collected in this study allows little more than

 

1George C. Homans, The Human Group (New York: Harcourt,
Brace and World, Inc., 1950). P. 112.

2Nathan Glazer and Daniel Moynihan, Beyond the Melting
Pot (Cambridge, Massachusetts: The M.I.T. Press, 1965),
p. 75.

4

a cursory examination of status differentials in interaction.
But I feel that this shortcoming does not necessarily lessen
the validity of the study. Homans argues that while the posi-
tive and negative quality and the structural aspects of con-
tact must not be disregarded in research, this "does not in-
validate our hypothesis, but does require that the influence
of these factors are stated in further hypotheses.“L

Taking this into account then, the central hypothesis
of this study can be stated:

Those who have had previous prolonged personal con-

tact with members of other races exhibit a higher

degree of racio-ethnic liberalism and social accept-

ance than do those who have had no previous prolonged

personal contact with members of other races.

The specific question operationalizing the contact vari-
able was--"Have you ever had prolonged personal contact with
people of other races?" If initial contacts between the mem-
bers of two races is perceived as negative on the part of one
or both parties, it is possible, of course, that further con-

tacts would be purposely avoided. Since the question asked

about prolonged personal contact, it is entirely possible

 

that those who answered that they had had no previous Egg:
longed contact did, indeed, have §9m§_contact with other
races. Thus, an answer of "no previous contact" could imply:
(1) If a small degree of contact was experienced, it may have
been perceived as negative--thus discouraging further inter-

action; (2) If a small degree of contact was experienced, and

 

1Homans, op. cit., p. 117.

it was perceived as being positive (or if no contact at all
was experienced), further or initial situational opportuni-
ties for interaction were not available or provided for
either or both sides; (5) Those who had no contact could have
had none because of previously held prejudices toward, or
stereotypes of, the other group--this would tend to nullify
motivation to seek interaction with other races or ethnic
groups.

Likewise, those who reported that they did have previous
prolonged personal contact with those of other races might
have experienced these contacts because: (1) They were previ-
ously racio-ethnically liberal and relatively unprejudiced so
that they at least did not resist, if not actively seek out,
contact with members of other races; (2) Their initial con—
tacts with people of other races or ethnic groups were per-
ceived as qualitatively so positive that they continued and
extended the contacts; or (5) The situational opportunities
for inter-racial contact were present, thus providing an en-
vironmental facilitation of interaction.

To say that previous prolonged personal contact between
the members of two racial groups is the significant factor
in the prejudiced or non-prejudiced attitudes of one group
toward the other (in this case, majority toward minority)
cannot be definitely established by the data gathered. Nor
can any hypothesis of attitude change be supported. That is,

no cause—effect relationship between the independent variable

of contact and the dependent variable of liberalism can be
established. Although there does seem to be a significant
relationship between the two variables, one cannot say
whether it is the factor of contact per se, or some inter—
vening variable, that produces the significance of the
relationship.
B. Configuration of Racio-ethnic
Prejudice or Ethnocentrism

A less central attempt of this study is to try to find
out just what types of people, in terms of the social and
demographic characteristics chosen to study, tend to be less
racio-ethnically liberal or more ethnocentric, regardless of
previous prolonged personal contact with members of other
races.

This concern is related to many studies which claim that
there is a certain personality type which can be identified
as being more prejudiced than other personality types, re-
gardless of the particular social object or social situation.
Bettelheim and Janowitz, for example, say that intimate con-
tact with minority group members does not seem to reduce
prejudice toward that group. It is not the contact that is
the significant factor, they argue, but rather the particular

personality structure of the party involved.1

 

lBruno Bettelheim and Morris Janowitz, Social Change and
Prejudice (Glencoe, Illinois: The Free Press, 1964), p. 72.

 

 

 

Himelhoch claims that the individual "will tend to ac-
cept and maintain the prejudiced pattern at the expense of
the democratic pattern if he has certain personality tenden-
cies for which prejudice provides an outlet and if other,
more satisfying, outlets are not provided."1 He presents
specific personality tendencies associated with high preju-
dice: status anxiety, authoritarian tendencies, repressed
sexual wishes, and insecurity arising from self—hate.2

Although this study is hardly prepared to delve into the
area of the "prejudiced psyche," it does try to identify some
cluster of social traits which shows a high correlation with
low racio-ethnic liberalism. It is not, of course, the

traits themselves that are significant, but rather what these

 

traits imply. For example, the finding that the females in
this study generally exhibit more prejudice than do the males
cannot be interpreted that simply being a female makes one

more ethnocentric or racio-ethnically prejudiced. Rather,

what is important is what being a female in our society means;
how the particular roles and statuses conventionally character-
izing the female help to form, and often dictate, her person-

ality and her relationships to others.

 

lJerome Himelhoch, "Is There a Bigot Personality?"
RaceLgPrejudice, and Discrimination, ed. Arnold Rose
INew York: Alfred Knopf, 1961), p. 500. This article origin-
ally appeared in Commentary, III (March, 1947), pp. 1277-84.

21bid.

C. Type of Contact

Although, as mentioned before, this study does not allow
a thorough examination of the differential impact of type of
contact, a few findings are presented which indicate to a
very preliminary degree that there might exist differences in
degree of racio-ethnic liberalism expressed on the basis of
whether the previous interaction was on a dominant-subordinate
level or on an equal status level. If the former can be in-
terpreted as a negative type of contact and the latter, as a
positive contact, one might expect those who had contact with
a minority group on an equal status level to be the most lib-
eral, those who had no previous contact to be the next most
liberal, and those who had engaged in dominant—subordinate
relationships to be the least racio-ethnically liberal or the

least socially accepting of out-racial groups.

III. METHODOLOGY
A. Operationalization of Variables

1. Contact--Independent Variable

Each respondent was asked to report his own past behavior.
The specific question was: "Have you ever had prolonged
personal contact with members of other races?“ The respondent
was to answer "yes" or "no", and "if yes, under what circum-
stances?"

After examination of the "circumstances" mentioned above,
responses were coded into three categories: (1) no, (2) yes,
contact with domestics, and (5) yes, contact with friends, co-
workers, classmates, etc. For my basic purposes, which deal
with existence of interaction rather than quality of inter-
action, categories (2) and (5) were combined so that the final
independent variable had two dimensions--“yes" and "no".

Also, a methodological difficulty arose in using the
three original categories; namely, that there were not enough
cases in category (2) yes, contact with domestics, to allow
statistically significant results after the application of
control variables. (However, a short presentation in percent-
ages of the data in the original three categories is presented,

later.)

10

11

2. Racio-ethnic Liberalism--Dependent Variable

The concept of racio—ethnic liberalism-~the dependent
variable--was operationalized by exacting responses to a
series of seven indicators or questions. It was intended that
the questions were worded in such a manner that, logically
speaking, an individual's response would be clearly indicative
of his position on a hypothetical scale of racio-ethnic liberal-
ism. In the final analysis, racio—ethnic liberalism was di-

chotomized into two categories, "high" and "low."

I

Each question is intended to provide a separate index of
the dependent variable; there is no composite "score" of all
the responses to form one index of racio—ethnic liberalism.
This is primarily to avoid the problem often encountered in
a Likert-type scale; namely, that a total "score" does not per-
mit an accurate analysis of what exactly is the response to
each question. Rather, the data gathered from each question
are discussed separately in perspective of the independent
variable.

However, there are what can be called "clusters" of the
seven indicators. For example, one cluster consists of the

1 Social Distance

three indicators that are Bogardus-type
Scales. These indicators call for the respondent to indicate
the degrees of social closeness to which he would permit a

member of the Negro, American Indian, and Oriental groups,

 

J'I say Bogardus-type because "to church as members" was
included between "to club as chums" and "to street as neigh-
bors."

12

respectively. In the Bogardus Scales, the respondent was
asked to indicate the highest, or closest, degree of his
acceptance of the members of each racial group. Thus, ideally,
indicating one category would imply that he would indicate

all lower, or less close, categories of social distance, and
no higher categories of social distance.

In the case of the Negro in this study, those who checked
"to kinship by marriage" as the closest level of acceptance
and those who checked "to club as personal chums" as the clos-
est level were combined into the group exhibiting high racio-
ethnic liberalism. All the others were combined to form the
low racio-ethnically liberal group.1

In both the cases of the American Indian and the Oriental,
those who indicated the highest level of acceptance possible--
"to kinship by marriage"--comprise the high liberal group; all
the other respondents comprise the low liberal group.

The second cluster of indicators is composed of the two
questions which deal slightly differently than Bogardus with
the closest level of social acceptance--marriage. In these,
the respondent was asked if, other things being equal, he
himself would marry someone of a different race, and if he
himself would marry someone of a different national background.
Original response possibilities were: (1) definitely yes,

(2) possibly, and (5) definitely no. On the question

 

1This combination of categories was necessary because
there were not enough cases who indicated the "to kinship by
marriage" level of acceptance to allow statistical presenta-
tion of data.

15

concerning race, those who answered "definitely yes" and.
"possibly" were combined1 and considered high in liberalism.
Those who responded "definitely no" were considered low in
liberalism.

On the other question—-concerning marriage with someone
of a different national background--those who answered
"definitely yes" were alone considered high in racio-ethnic
liberalism; those who answered "possibly" and "definitely no"
were combined to form the category of low racio-ethnic liberal-
ism. (Aside from methodological considerations, this collapse
of categories is due to the probability that national back-
ground is not as stringent a test of racio-ethnic liberalism
as is race.)

The last two of seven questions or indicators compose
the third cluster, which can be termed the "legal" aspect of
racio-ethnic liberalism. This is to be contrasted with the
above five indicators which compose the "social" aspect of
racio-ethnic liberalism. These last two questions asked the
respondent to express his attitude (selecting from the alterna—
tive responses presented) about a law affecting relationships
between the majority and minority groups in this country.

The first indicator asked the respondent if he was in accord
with the 1954 Supreme Court decision ordering racial integra-

tion of all schools. Those who responded "strongly in

 

1This combination was partially due to the lack of suf-
ficient cases who answered “definitely yes."

14

agreement" and "somewhat in agreement" were combined and
termed high in racio—ethnic liberalism. Those who responded
"undecided," "somewhat in disagreement," and "strongly in
disagreement" were combined and termed low in racio-ethnic
liberalism.

The second "legal" indicator asked the respondent, after
an explanation of anti-miscegenation laws, if he thinks that
these laws are unconstitutional. Those who answered "very
definitely unconstitutional" and "mostly unconstitutional"
were grouped to form the high liberal category; those who re-
sponded "undecided," "mostly constitutional" and "very defi—

nitely constitutional" together compose the low liberal cate-

gory.

5. Control Variables

Of the socialanxidemographic characteristics obtained
from the population, seven which seemed most relevant were
chosen and applied as controls in testing the general hypothe-
sis. Due to lack of sufficient cases, use of simultaneous
controls was not_possible. The control variables are: sex,
college status, size of home town (the one in which the re-
spondent spent most of his first sixteen years), frequency
of church attendance, status of father's occupation, grade
point average, and general political orientation.

In the case of college status, respondents were asked to

indicate (1) freshman, (2) sophomore, (5) junior, (4) senior,

 

15

or (5) graduate student. Two final categories were formed
from this--freshman and non-freshman.

The size of home town question provided five alternative
responses: (1) farm and country, (2) small town (under 2,500),
(5) small city (2,500-25,000), (4) large city (25,000—

100,000) and (5) metropolis (over 100,000). The cases of the
first three were combined to form a small town category, and
the cases of the last two compose the large town category.

The question of frequency of one's church attendance

 

allowed five possible responses originally: '(1) at"least once
a week, (2) at least twice a month, (5) about once a month,

(4) a few times a year, and (5) do not attend. The first

three response groups were combined to form the high frequency
of church attendance category, and the last two response groups
together comprise the low frequency of church attendance
category.

Each respondent was asked to briefly state his father's
occupation. The standard by which status was attributed to
each One is the North Hatt Scale of Occupations. On the basis
of one's occupational score according to the North-Hatt Scale
(from 50-59, 40—49, 50-59, 60-69, 70 and over), each occupa-
tion was placed in one of the five status categories. After
this, responses from the two highest status levels were com-
bined to form the high status category; the third highest status
level itself serves as the middle status category; the two

lowest status levels together compose the low status category.

 

16

The original questionnaire provided seven categories of
grade-point responses: (1) 4.0, (2) 5.5, (5) 5.0, (4) 2.5,
(5) 2.0, (6) 1.5, and (7) below 1.5. In the final analysis,
4.0 and 5.5 comprise a high grade point category, 5.0 and
2.5 comprise a medium grade point category, and 2.0, 1.5 and
below 1.5 together form the low grade point category.

General political orientation was ascertained by having
the reapondents identify themselves with one of the following:
(1) quite liberal, (2) liberal, (5) moderate, (4) conservative,
and (5) quite conservative. The "quite liberal" and "liberal"
responses were combined to form the liberal political group,
the "moderate" responses themselves form the moderate political
group, and the "conservative“ and "quite conservative" cases
together comprise the conservative political group.

The basis for collapsing the cases from the original re-
sponse categories to those more general categories by which
the data is herein analyzed, is twofoldehmethodological and
theoretical. Methodologically, the lack of sufficient cases
in some of the original response categories necessitated some
maneuvering to attempt to offset possible distortion of re—
sults. Theoretically, some categories just more logically
clustered together than others--for example, "quite liberal"
and "liberal." But, at times, when neither methodological or
logical considerations were seriously involved, it was, ad-

mittedly, a relatively arbitrary decision.

17

B. Research Population and Data Collection

The sample used in this study was chosen on a non-random
basis. The specific research population consisted of 265
white college students, all members of general sociology
classes. Data were collected at the University of Missouri
in Columbia during a three—week period in April, 1965.

The method of data collection was the questionnaire.

Since the sample consisted entirely of students, the question-

 

naires were administered in group form (during class periods);
each class consisted of approximately fifty students. The
great majority of questions provided the alternative response
categories from which the respondent had to choose. Some
"open-ended" questions were used. These responses were content
analyzed, and after an apparently valid configuration of re-
sponses was obtained, responses were coded categorically.

For this study, the questionnaire was preferred to other
methods of data collection, principally because of the anony-
mity that a questionnaire allows. The nature of many of the
questions was such that a respondent might be reluctant to
answer honestly in a face—to—face, interviewer-interviewee
relationship.

The chi—square method of analysis was applied to the data
where concern was with impact of contact on attitudes (or
association between contact and attitudes). Other problems
dealt with in this paper--characteristics associated with
prejudice, and the status differentiation of an intergroup

relationship——are presented as simple percentages.

IV . FINDINGS

A. Association Between Previous Contact
and Racio-ethnic Liberalism

As is shown in Table I for the three Bogardus Scale indi-
cators, the relationship between previous personal prolonged
contact and racio-ethnic liberalism is significant (at the
.05 level or below) positively and directly as hypothesized
for certain groups.

It can be seen that the relationship is significant for:
females more often than males; for freshmen and non-freshmen
alike; in all cases for those from a small town, and in no
cases for those from a large town; for those having a low
grade-point in all three cases and not significant in any case
for those having medium and high grade points; in more in-
stances for those attending church with high frequency than
for those attending church with low frequency; more often for
those with a moderate political orientation than for those
with a conservative political orientation, and in no case for
political liberals; and in all cases for those whose fathers
are in the middle status occupations, followed by those fathers
in the high status occupations, and in no case for those
fathers in the low status category.

In terms of the three indicators, the relationship pre-

dicted between contact and liberalism is significant most

18

No.

No.

No.

No.

No.

No.

No.

19

FIGURE 1

CODE FOR INDICATORS OF QUESTIONS

1--Bogardus Social Distance Scale—-Negro
2--Bogardus Social Distance Scale--American Indian
5--Bogardus Social Distance Scale--Oriental
4-4Would marry someone of a different race

5-—Would marry someone of a different national back-
ground

6--Attitude toward 1954 Supreme Court decision re
school integration .

7—-Attitude toward constitutionality of anti—
miscegenation laws ‘

 

20

Table I. Chi—square Significance of Predicted Relationship
Between Contact and Liberalism

 

 

 

 

 

 

 

 

Low status

No. 2
No. 1 * Bogardus- No. 5
Bogardus- American Bogardus-
Negro Indian Oriental
Sex
Male .50 .10 .01
Female .001 .20 .05
College Status
Freshman .05 .10 .05
Non-freshman .01 .20 .02
Size of Home Town
Small .001 .001 .001
Large .50 .99 .20
Grade-Point
High .20 .50 .20
Medium .20 .70 .50
Low .001 .01 .001
Church Attendance
High frequency .01 .10 .02
Low frequency .50 .20 .05
Political Orientation
Liberal .10 .50 .20
Moderate .05 .20 .01
Conservative .20 .01 .10
Father's Occupation
High status .10 .50 .01
Middle status .05 .01 .02

.10 .99 .50

 

 

21

often for the Scale concerning the Oriental group, next often
concerning the Negro group, and least often significant cone
cerning the American Indian group.

The relationship is most consistently significant for
those of a small town, those having a low grade point, and
those whose fathers' occupations fall in the medium status
group.

For the next cluster of indicators, which deals with

projected personal acceptance of a spouse of a different racial

 

and different national group, respectively, the findings are
presented in Table II.

Here, contact is a significant factor: more often for
females than males; in both cases for freshmen and for non-
freshmen; in both cases for those from both small and large
town backgrounds; in both cases for those with a medium grade-
point and in one case for both those with high and with low
grade points; more often for those who attend church with a
high frequency than for those who attend with a low frequency;
in both cases for political liberals and moderates and in
neither case for political conservatives; and in both cases
for the high and middle occupational groups, but in neither
case for the low occupational group.

It can be seen that of the two indicators, the one con-
cerning race shows a greater number of significant relation-
ships (15) between the independent and dependent variable.

In fact, for all control categories, the relationship is

22

Table II. Chiesquare Significance of Predicted Relationship
Between Contact and Liberalism

-_
_

 

 

 

 

 

 

 

 

No. 4 No. 5 "
Marry of Marry of differ-
different ent national
race background
Sex
Male .02 .10
Female .001 .01
College Status
Freshman .001 .01
Non-freshman .01 .02
Size of Home Town
Small .001 .01
Large .02 .02
Grade Point
High .02 .10
Medium .05 .05
Low .01 ' .20
Church Attendance
High frequency .001 .01
Low frequency .05 .10
Political Orientation
Liberal .05 .05
Moderate .01 .01
Conservative .20 .10
Father's Occupation
High status .001 .01
Middle status .01 .01

Low status .20 .50

 

25

significant except for two--political conservatives and the
low status occupational group.

In Table III the "legal" cluster of two indicators shows
a significant relationship between contact and liberalism:
more often for females than males; more often for non-fresh-
men than freshmen; more often for those from a small town
than a large town; most often for those with low grade points,
next often for high grade pointers, and in neither case for
those with medium grade points; more often for high church
attenders than for low church attenders; in both cases for
political liberals and in neither case for moderates or con-
servatives; most often for the low occupational group, next
most often for the middle occupational group, and not at all
for those in the high occupational group.

Of the two legal indicators in question, and one concern—
ing the 1954 Supreme Court decision (no. 6) presents a larger
number of significant relationships (11) than does the one
concerning the constitutionality of anti—miscegenation laws
(9).

Looking at the findings for all indicators together in
terms of each control group, a pattern seems to emerge.

A significant relationship between contact and attitude ap—
pears more often for females than males; more often for non—
freshmen than freshmen; more often for those of a small town
than for those of a large town; most often for those with a

low grade point; more often for those with a high church

24

Table III. Chi-square Significance of Predicted Relationship
Between Contact and Liberalism

 

 

 

 

 

 

 

 

No. 6 No. 7
Supreme Anti—
Court miscegenation
decision laws
Sex
Male .01 .10
Female .05 .01
College Status
Freshman .10 .05
Non—freshman .001 .01
Size of Home Town
Small .02 .001
Large .05 .10
Grade Point
High .50 .02
Medium .20 .50
Low .001 .02
Church Attendance
High frequency .02 .01
Low frequency .05 .10
Political Orientation
Liberal .02 .001
Moderate .50 .50
Conservative .20 .50
Father's Occupation
High status .50 .20
Middle status .02 .20

Low status .05 .01

 

25

frequency than those with a low church frequency; equally as
often for both political moderates and liberals (but both

more often than political conservatives); and most often

for those whose fathers' occupations fall in the middle status

Q

category.

B. Identification of Some Correlates
of Racio-ethnic Prejudice

Three out of the total of seven indicators were chosen
as representative for the purpose of ascertaining which of
the categories within each control group is the more liberal.

Question number 1 was chosen from the three Social
Distance Scales, because one's degree of acceptance of Negroes
would seem to be the most discriminating indicator of racio—
ethnic liberalism.

Question number 4 represents the cluster of two questions
which calls for indicating one's probable behavior in a
marriage situation, because again, race is probably a more
discriminating factor for most people than national back-
ground.1

Question number 6 was chosen from the two "legal" indi—
catorsausbeing the one that dealt with an issue to which the
general population has been much more thoroughly exposed;

thus, respondents would more likely hold a clear attitude

 

1It can be pointed out here that this question--number
4-—proved to be the most often statistically significant of
all the seven indicators used.

 

26

toward it. One might also interpret both legal questions as

representing the more formal, perhaps even superficial, as-

pects of racial attitudes. Findings are presented in simple

percentages rather than chi-squares because this particular

concern is not central to the study. Rather, these data, by

indicating trends and patterns, are intended only to supple-
ment the basic problem of the study--namely, the relationship

between contact and attitudes.

 

 

 

 

 

 

 

 

 

 

 

 

 

1. Sex
No. 1 No. 4
Racio-ethnic Racio-ethnic
Liberalism Liberalism
Hi Low Hi Low .
(62) (40) . (52) (55)
male male
60.8 59.2 49.5 50.5
(86) (68) (51) (105)
female female
55.8 44.2 55.1 66.9
No. 6
R-E Liberalisml
Hi Low
(85) (25)
male
78.5 21.7
(126) (30)
female
80.8 19.2

 

 

 

 

 

 

 

ethnic Liberalism.

1Hereafter, R-E Liberalism will be used to mean Racio-

27

The males are more liberal than the females on indicators

numbers 1 and 4. The females are more liberal on the legal

indicator number 6.

2. College Status

No. 1 No. 4

R-E Liberalism

R?E Liberalism

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Hi Low Hi Low
(99) (70) (68) (104)
freshman freshman
58.6 41.4 59.5 60.5
non_ (50) (57) non_ (36) (51)
freshman 57. 42.5 freshman 41.4 58.6
No. 6
R-E Liberalism
Hi Low
(158) (56)
freshman
79.5 20.7
non- (71) (18)
freshman 79.8 20.2

 

 

 

 

Differences between freshmen and those above freshman

status are small.

greater degree of liberality;

reversed--non-freshmen are more liberal.

On Question number 1,

on Questions 4 and 6,

the freshmen show a

it is

28

5. Size of Home Town
No. 1 No. 4 No. 6
R—E Liberalism

R-E Liberalism R-E Liberalism

 

 

 

 

 

 

Hi Low Hi Low Hi Low
(70) (56) (50) (77) (101) (27)

small small small
55.6 44.4 59.4 60.6 78.9 21.1
(76) (55) (55) (77) (108) (26)

large large large
58.9 41.1 41.7 58.5 80.6 19.4

 

 

 

 

 

 

 

 

 

 

 

 

Again, differences between those of a small town background

 

and those of a large town background appear slight, but in all

three cases, those from the larger towns are more liberal.

4. Frequency of Church Attendance

No. 1 No. 4

R-E Liberalism R-E Liberalism

 

 

 

 

 

 

 

 

 

 

 

 

Hi Low Hi Low

high (95) (71) high (66)L (102)

frequency 57.2 42.8 frequency 59.3 50,7

low (55) (55) low (58) (51)

frequency 59.6 40,4 frequency 42.7 57.5
No. 6

R-E Liberalism

 

 

Hi Low
high (159) (52)
frequency 81.5 18.7
low (69) (21)
frequency 76.7 25.5

 

 

 

 

29

On Questions numbers 1 and 4, low church attenders ap—
pear to be more liberal--on Question 6, the high attenders

evidence a greater degree of liberalism.

5. Status of Father's Occupation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

No. 1 , No. 4
R-E Liberalism R-E Liberalism
Rank Hi Low Rank Hi Low
high (26) (25) high (20) (29)
5 1

Status 55.1 46.9 Status 40.8 59.2
2 middle (69) (56) 3 middle (47) (75)
Status 55.2 44.8 Status 58.5 61.5
low <47) (27) low (51) (44)

1 status 2 status
65.5 56.5 41.5 58.7

No. 6

R-E Liberalism

 

 

 

Rank Hi Low
Status 87.8 12.2

3 middle (96) (50)
Status 76.2 25.8

1 low (61) (14)
status 81.5 18.7

 

 

 

There does not seem to be a consistent pattern in the
relationship between the status of father's occupation and
liberalism. On both Questions numbers 1 and 4, those of low

status evidence more liberal attitudes than either the middle

50

or high status occupational groups. On Questions numbers 4
and 6, the middle status group shows the least degree of
liberalism. If we rank--from highest degree of liberalism to
lowest degree of liberalism--each group on each question, we
can obtain a composite "score" of rank numbers. So, for high
status the score is 6, middle status is 8, and low status, 4.
For the combination of questions, then, the middle status oc-
cupational group is the least liberal, followed by the high

status and low status groups, respectively.

6. Grade Point

 

 

 

 

 

 

 

 

 

 

 

 

 

 

No. 1 No. 4
R-E Liberalism R—E Liberalism

Rank Hi Low Rank . Hi Low
high (46) (17) high (27) (57)

1 grade 1 grade
point 75.0 27.0 point 42.2 57.8
medium (77) (59) medium (52) (85)

2 grade 5 grade
point 56.6 45.4 P01nt 58.5 61.5
low (26) (51) low (24) (54)

5 grade 2 grade
point 45.6 54.4 point 41.4 58.6

No. 6

R—E Liberalism

Rank Hi Low
high (54) (11)

1 grade
point 85.1 16.9

medium (107)_ (28)

 

 

 

2 grade
point 79.5 20.7
low (47) (15)
5 grade

 

 

 

point 78.5 21.7

 

 

51

In terms of one's grade point average, those with high
grade points are more liberal on every question than the
other two categories-—medium and low grade point. The mediums
are more liberal than the lows on Questions numbers 1 and 6;
it is reversed on Question 4. Again, in terms of rank on
the three questions, it is found that the high grade pointers

are most liberal, the medium grade pointers the next most

liberal, and the low grade pointers the least liberal.

7. Political Orientation

No. 1 No. 4

R—E Liberalism R-E Liberalism

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Rank Hi Low Rank Hi Low
(57) (44) - (45) (55)
2 liberal 1 liberal
56.4 45.6 45.0 55.0
(65) (58) (59) (62)
1 moderate 2 moderate
65.1 56.9 58.6 61.4
(25) (27) (17) (55)
5 conservative 5 conservative
46.0 54.0 52.7 67.5
No. 6
R-E Liberalism
Rank Hi Low
(86) .(16)
1 liberal
84.5 15.7
(87) (17)
2 moderate
85.7 16.5
(55) (19)
5 conservative
65.5 56.5

 

 

 

 

52

In all three cases, those with a conservative political
orientation are the least liberal. In two of the three cases,
the political liberals are more racio—ethnically liberal than
the moderates. In one case (Question number 1), the moderates
are more racio—ethnically liberal than the political liberals.
When we rank the three political categories in terms of a
composite score, we can see that the political liberals are
the most "accepting," the political moderates next, and the
political conservatives the least "accepting."

All of the above findings are not intended, obviously,
to positively identify those characteristics that cause one
to exhibit more or less liberal attitudes toward other races.
Nor do they intend to identify characteristics that must be
present in order that a person express either liberal or pre—

judiced attitudes. The findings are meant to indicate the

social and demographic variables which, when in interaction

 

with other variables, might tend to be associated with certain
types of attitudes. On the basis of this, then, it appears
that a configuration of racio-ethnic prejudice or ethnocentrism
consists, in this study, of females, freshmen, those from a
small town, those who attend church with a high frequency,
those whose fathers are of the middle occupational status
group, those who have low grade points, and political conserva—

tives.

 

55

C. Type of Contact

The following tables present the results of running type

of previous contact against each indicator of racio—ethnic

liberalism.

no
previous
contact yes,

with domestics

yes,

with peers1

No. 1

R-E-Liberalism

 

 

 

 

 

 

 

Hi Low
(69) (73)
48.6 51.4
(8) (7)
55.5 . 46.7
(48) (50)
71.4 28.6

No. 5

R-E Liberalism

 

 

 

 

 

 

 

No. 2

R-E Liberalism

 

 

 

 

 

 

 

Hi Low
(51) (90) .
no '
56.2 65.8 (
previous 5.
contact yes, (5) (10) '
With domestics 53-5- 66.7
yes, (54) (45)
With peers 54.6 45.5
No. 4

R-E Liberalism

 

 

 

 

 

 

 

 

Hi Low Hi Low
(58) (104) (40) (102)
no no
26.8 75.2 28.2 71.8
previous previous
contact yes, (6) (10) contact yes, (5) (11)
With domestics 57.5 62.5 Wlth domestics 51.5 68.8
yes, (48) (50) yes, (58) (41)
with peers with peers
49.0 51.0 58.6 41.4
lPeers includes classmates, co-workers, friends, etc.

54

No. 5 No. 6

R-E Liberalism R—E Liberalism

 

 

 

 

 

 

 

 

 

 

 

 

 

 

On five of the seven indicators there

 

 

 

 

 

 

Hi Low Hi Low
(46) (97) (105) (40)
no no
52.2 67.8 72.4 27.6
previous (6) (10) previous (15) (1)
contact yes, contact yes,
with domestics 57.5 62.5 with domestics 95.8 6.2
yes, (57) (45) yes, (88) (12)
with peers with peers
57.0 45.0 88.0 12.0
No. 7
R—E Liberalism
Hi Low
(96) (49)
no
. 66.2 55.8
preVious
contaCt yes, (11) (5)
With domestics 68.8 51.2
(87) (15)
yes,
with peers 87.0 15.0

is the same rela-

tionship pattern between type of contact and the dependent

variable of racio-ethnic liberalism.

That is,

experienced equal status contacts are the most liberal,

those who have

those

who have experienced dominant-subordinate relationships (with

domestics, primarily) are the next most liberal, and those

Who have had pp previous prolonged contact are the least

liberal.

 

 

55

On Question number 6, those experiencing dominant—sub—
ordinate relationships appear more liberal than those
experiencing equal status contacts.

The other exception to the general pattern is Question
number 2-—the Bogardus Scale concerning the American Indian.
The "no" contacts are more liberal than the contacts "with
domestics," but are still less liberal than the contacts
"with peers."

Interpretation of all the above data must be cautious,
due to the small number of cases in the dominant-subordinate
(domestic) category. A relatively large percentage shift

would result from a shift of only one or two cases.

V. SUMMARY AND DISCUSSION

Females appear to be less racioeethnically liberal than
do males, and contact is more often significantly related to
racio-ethnic liberalism (in the manner hypothesized) for fe-
males than for males. (

Perhaps that women evidence less liberalism, at least in

terms of social distance, is a function of the role of the

4‘ ." -23 _

woman in our culture. The female is more often bound by
social conventions, and is expected to be more conservative
than the malejxlheterosexual relationships. Relationships
with a man of another race may be more potentially threatening
to a woman than a man. Also, especially in the South, the
preservation of "white womanhood" is symbolic of the mainte—
nance of caste relations between the races.

The fact that contact is more often significantly related
to liberalism for women than men could also be related to the
idea that a woman is more often socially isolated, less mobile,
and has less access to various types of contact. Thus, she
is allowed to build up a strong system of prejudice. So when
she does experience interracial contact, it is more likely to
effect a stronger reduction of prejudice than it would for

men .

56

57

The slightly higher liberality expressed by Women on
Question number 6--concerning the Supreme Court decision--
suggests a tendency to conform to formal, institutional and
less personal proscriptions.

Differences in college status are relatively slight, but
those above freshman status appear more racio—ethnically
liberal than freshmen. However, the contact variable has a
slightly greater impact on the attitudes of non—freshmen than
freshmen.

It might be suggested here that those above freshman
status are a bit more sophisticated in outlook.1 For many
students, college is the first setting in which they interact,
or are provided possibilities to interact, with people of
other races and nationalities on a peer basis. Perhaps when
the non—freshmen in this sample were themselves freshmen, they
held less liberal attitudes than they do now. Thus, due to
increased exposure over time, one might expect that for fresh-
men, contact would have a more significant impact in reducing
previously held, and perhaps untried, prejudices. But the
data do not bear out this expectation.

Those who come from larger towns are more racio-ethnically

liberal than those from smaller towns. Also, contact and

 

1For a study relevant to this particular concern, see
Newcomb's study of Bennington College women. Theodore M.
Newcomb, "Some Patterned Consequences of Membership in a Col-
lege Community," Readings in Social Psychology, ed. Theodore
Newcomb and Eugene Hartley (New York: Henry Holt and Company,
1947), pp. 545-557.

58

liberalism are more often significantly related as hypothe-
sized for small town students than for large town students.

That those from a large town background demonstrate a
higher degree of racio-ethnic liberalism suggests that environ—
mental factors might be operative here. The more hetero—
geneous, secular and anonymous nature of a large city is more
likely to provide available channels for various types of in-
teraction——even interaction that appears to go against the
norms. As well, a general atmosphere which necessarily toler—
ates, if not levels, differences between and among people,
offers a type of social sanction to interracial relationships.
It is likely that social controls upon individual behavior
are less rigidly enforced than they are in a small town.

When controls are often more of a secondary than a primary
nature, the pressures from the others in one's in—group against
out-group contacts are likely to be less strong in the large
city than the small town.

In most cases, those with a low frequency of church
attendance are more racio-ethnically liberal, regardless of
contact, than those with a high frequency of church attendance.
Likewise, the relationship between contact and liberalism is
significant twice as often for high church attenders as for
low church attenders.

A suggestion as to Why high church attenders appear
slightly less liberal is that most often the church is a de—

fender of the status quo and an upholder of traditional norms

 

59

and behavior patterns. Thus, the church might implicitly dis;
courage such "unconventional" behavior as interracial con-
tacts.

Personality theories of prejudice might contend that high
church attenders are more doctrinaire and exhibit stronger
traits of dogmatism. Thus, one would expect them to maintain
Wider social distances between themselves and members of other
races and ethnic groups. But when it comes to the less per—
sonal aspectscflfracial acceptance--as in Question number 6
concerning the Supreme Court decision-~racio-ethnic liberalism
is higher for them because, perhaps, supporters of one insti-
tutional structure might be more likely to support another
institution, which, in this case, is the legal system.

We can conjecture that because of a certain social iso—
lation (or social insularity) of high church attenders, inter-
racial contact would have more of an effect upon their atti-
tudes than upon the attitudes of low church attenders.

For the occupational variable, regardless of contact, the
low status occupations generally show the highest degree of
racio-ethnic liberalism, followed by the high status group
and middle status group, respectively. When contact is taken
into account, it is a significant factor in attitudes most
often for the middle status group, next often for the high
status group and least often fOr the low status group.

In the search for explanations of these findings, it must

first be realized that the status of the occupation of one's

40

father is not necessarily the status of the respondent him—
self. The sample consists entirely of college students; it
is likely that the respondent's future socio—economic status
will fall above that of his father's, especially if the latter
falls in the low status occupational group. Since these re-
spondents could be called upwardly mobile, and some studies
have found upwardly mobile people more tolerant than others,
one could expect the low status respondents to appear the most
tolerant.l

Those whose fathers' occupations fall in the middle status
group show the least tolerance, regardless of Contact, and
also show the greatest number of significant relationships
(in the manner hypothesized) between contact and liberalism.

If the middle status group may be considered the most
strongly dependent upon the status quo and therefore, less
favorable to anything that might represent change of that
status quo, then it seems logical that they would express such
a low degree of liberalism toward other races.

In regard to the contact variable, it might be that of
the three groups, those of middle class backgrounds are the
most sheltered from interracial contacts. The high status

people more likely have access to relationships with Negro

 

1For a discussion of how upward mobility and status in—
consistency are related to liberal attitudes, see Gerhard E.
Lenski, "Status Crystallization: A Non—Vertical Dimension
of Social Status," American Sociological Review, XIX (August,
1954), pp. 405-415.

41

professionals as well as employees of other races. Low sta-
tus people might have more contact by reasons of residential,
educational, and occupational proximity. The middle class,
especially if they live on the fringe areas of Negro neigh-
borhoods, are likely to look upon interracial contacts as
threatening: socially, economically (property values), and
politically. Assuming that the middle class is most often
the least secure, then, jeopardizing contacts of any kind
would be strongly avoided.

Those with the highest grade point appear more liberal
than either the medium grade point or low grade point groups.
And in most cases, the medium group is more liberal than the
low grade point group.

It was also found that previous personal contact varies
directly with racio-ethnic liberalism at a significant level
most often for the low grade point group.

Although we cannot presume that grade point average is
indicative of intelligence, we might be able to say that those
with a higher grade point have demonstrated, by some standard
of appraisal, a firmer grasp of "matters of the mind." Hence,
they might possess more knowledge of people of other races
and ethnic groups which Allport, for one, considers an impor-
tant factor in reduction or prejudice.1

Since it is for the low grade point group that contact

and liberalism are most often significantly related, one is

 

lAllport, op. cit., p. 254.

42

tempted to speculate that both the high and medium grade point
groups possess a certain intellectual awareness which could
predispose them favorably to another racial or ethnic group.
That is, personal contact may not be as important for them;

it might only give emotional substantiation to what rational
beliefs they already hold. But maybe those with low grade
points require a direct confrontation with the object of their
prejudice in order to undergo any degree of attitude change.

Regardless of contact, the political conservatives ex—
hibit the least degree of racio—ethnic liberalism, and the
liberals exhibit a higher degree of racio-ethnic liberalism
than do the moderates on two of the three representative
indicators.

The predicted relationship between contact and liberal-
ism is significant as frequently for liberals as for moder-
ates, and least often significant for conservatives.

That the conservatives are the least liberal would be
reasonable if we assume that political conservatism varies
directly with conservatism in the social realm. By the same
assumption, it is likely that the political liberals would
exhibit a greater degree of social liberalism toward other
races and ethnic groups than either the moderates or the con-
servatives.

We would expect contact to be significant regarding
liberal attitudes more often for the moderate group than for

either the liberals or conservatives. The reason for this

45

expectation might be that the moderates, by the very nature
of their own political identification, hold less intense,
if not less extreme, attitudes on a variety of issues.
From this, both liberals and conservatives would be more
resistant to attitude change than the moderates-—in this case,
attitude change through the process of interracial contact.
But the findings show the relationship significant frequently
as often for moderates and liberals.

There appear to be definite similarities between impact
of contact and the correlates of racio-ethnic prejudice.
To review briefly, those appearing to be least racio-ethnically
liberal, regardless of contact, are females, freshmen, small
towners, low grade pointers, political conservatives, and
fathers in the middle status occupations. Those for whom the
variable of previous prolonged personal contact and the vari—
able of racio-ethnic liberalism are most often significantly
related in the manner hypothesized are, with two exceptions,
the same groups of people. The two exceptions are college
status, where the relationship is slightly reversed, and
political orientation, Where the number of significant cases
is more for both liberals and moderates‘than conservatives.

That these similarities appear as they do suggests a
further specification of the relationship predicted in the
general hypothesis. That is, in cases Where people previously
hold strong prejudices, personal contact with the social ob-

jects of their prejudices results in a significant reduction

44

of same. It follows, then, that those persons who hold rela;
tively weaker prejudices seek out, or at least do not resist,
contact with other races. Thus, these people would tend to
exhibit a relatively low amount of prejudice both before and
after interracial contact——or regardless of contact.

Although this study is not prepared to substantiate this
statistically, it does suggest the following hypothesis for
further research in this area: Interracial contact serves to
significantly reduceyprejudice if the person of the majority
group_already holds strong_prejudices, but not so for those
who previously possess weaker_prejudices. Further research

should deal with the perceived nature of contacts and the

 

effectiveness of various types of contacts. It is possible,
for example, that negatively perceived contacts serve to
strengthen previously held prejudices, and that positively
perceived contacts serve to reduce previously held prejudices.

For people relatively free of prejudice before contact, per—

 

ceived nature of contact could be expected to have less of an
impact.

When the pypg of contact is considered, it is found that
in most of the cases (indicators), those who have had contacts
with peers are the most liberal, those who have had contacts
with domestics are the next most liberal, and those who have
had no contacts at all with other races are the least racio—

ethnically liberal.

45

One exception to this pattern is Question number 2--the

Bogardus Scale concerning the American Indian—-on which the

no contacts are more liberal than the contacts "with domes-
tics."

It might be relevant to note that the American Indian
is probably the least threatening racial minority group today.
It is the least threatening in various aspects: 1) Physically;
their characteristics do not immediately stand out from the
rest of the White population--their visibility is relatively
low. 2) Ecologically; they come into contact with a relatively
small proportion of White Americans due to regional concen-
tration. 5) Politically; there appears to be little, if any,
organized protest or revolt against the status quo or white
power structure--at least compared to the Negroes.

It should be remembered that the very small number of
cases in the contact "with domestics" category could produce
distortion in the over-all relationships among the three cate—
gories of contact. Putting this methodological problem aside,
however, one can suggest that these figures might be indica—
tive of a more general trend in interracial/ethnic relations.
That is, ppmp prolonged personal contact, regardless of the
status differential that might be involved, could predispose

one to be more racio-ethnically liberal than if there were no

previous prolonged personal contact at all.

VI. CONCLUSION

Among the many ways in which this study is circumscribed
are the following:

(1) The findings apply only to the sample in this study.
Since college students compose a rather specific segment of
the general population, findings are not generalizable to the
rest of the population or to other specific segments within
society.

(2) Conceptual problems arise from the nature of the
sample. For example, it may be untenable to assume that an
18 year old student is aware of what his political orientation
is.

(5) This research has carried the implicit assumption
that verbal behavior is a fair measure of "actual" behavior.
That is, asking a respondent what he would do in a certain
situation does not necessarily mean that, when actually in that
social situation, he would perform as he said he would.

(4) The manner in which the original response categories
were collapsed may have led to a less discriminating and more
distorted analysis. For example, those with a 2.0 grade point
were placed in the low grade point category and those from
towns of 25,000 were placed in the small town category.

Clearly, these are rather arbitrary decisions. If the sample

46

47

had consisted of a greater number of cases, these combinations
of categories might not have been necessary. Similarly, a
greater number of cases would have allowed for the applica-
tion of simultaneous controls, which might have produced dif-
ferent results.

(5) This study did not intend to deal with all the social
and demographic factors that might have a bearing on the
general hypothesis. It is, of course, possible that some
crucial factors were excluded.

(6) The seven indicators used were not subjected to statis-
tical analysis to determine how well they tapped the independent
variable of racio-ethnic liberalism. The indicators were
chosen on the basis of their face validity, and the primary
emphasis was upon tentative patterns, rather than statistically
demonstrated particular relationships.

(7) The criteria of the North-Hatt Scale can be easily
debated. For instance, a farmer was placed in the middle
status category along with some types of white collar workers.
This could have distorted the data when generalizing about
occupational groupings.

While the limitations are serious, there is warrant to
see within the data a case for further research. The central
objective of this study was not to establish cause-effect
relationships, but rather to describe and suggest. In that
context, the data may support a guarded optimism that can be

carried into a more sophisticated experimental framework.

BIBLIOGRAPHY

Allport, Gordon W. The Nature of Prejudice. New York:
Doubleday and Co., 1966.

Bettelheim, Bruno and Janowitz, Morris. Social Change and
Prejudice. Glencoe, Illinois: The Free Press, 1964.

 

Brophy, Ira N. "The Luxury of Anti—Negro Prejudice,"
Public Opinion Quarterly,‘IX (Winter, 1945—46), 456—466.

 

Deutsch, Morton and Collins, Mary E. Interracial Housing.
Minneapolis, Minnesota: University of Minnesota Press,
1951.

Glazer, Nathan and Moynihan, Daniel P. Beyond the Melting
Pot. Cambridge: The M.I.T. Press, 1965.

 

Himelhoch, Jerome. "Is There a Bigot Personality?" in
Rose, Arnold M. (ed.) Race, Prejudice and Discrimination.
New York: Alfred A. Knopf, 1951, pp. 487-501.

 

Homans, George C. The Human Group. New York: Harcourt,
Brace and World, Inc., 1950.

 

Lenski, Gerhard E. "Status Crystallization: A Non-vertical
Dimension of Social Status," American Sociological
Review, XIX (August, 1954), 405—415.

 

Myrdal, Gunnar. An American Dilemma. New York: Harper
and Brothers Publishers, 1944.

 

Newcomb, Theodore M. "Some Patterned Consequences of Member-
ship in a College Community," in Readings in Social
Psychology. Theodore Newcomb and Eugene Hartley (ed.).
New York: Henry Holt and Company, 1947, 545—557.

 

Rose, Arnold M., ed. Studies in Reduction of Prejudice.
Chicago, Illinois: American Council on Race Relations,
1948.

Simpson, George E. and Yinger, J. Milton. Racial and Cultural
Minorities. New York: Harper and Row, 1965.

 

Williams, Robin M., Jr. Strangers Next Door. Englewood
Cliffs, New Jersey: Prentice-Hall, Inc., 1964.

 

48

l
l
l
l
l
l
l
l
l
I'
l
l
l
I
I
l
l
i
l
l
l

9
2
1
3