A TEST OF S'TE'REOTYPE ACCURACY

Thesis Ior “)9 Dear” of M. A,
MICHIGAN STATE UNIVERSITY

Albert Zavala
1960

 

 

 

 

 

A TEST OF STEREOTYPE ACCURACY

by

Albert Zavala

AN ABSTRACT

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

MASTER OF ARTS

Department of Psychology

1960

Approvek LQQ am
«A

ABSTRACT
A TEST OF STEREOTYPE ACCURACY
by Albert Zavala

The purpose of this study was to develop an objective and
economical test that would measure Stereotype Accuracy, or the ability
to predict the behavior of social groups.

Four 30-item scales were designed to measure the accuracy
of four stereotypes: College Man Stereotype, Woman Stereotype, Sex
Stereotype, and Age Stereotype. These were developed from four
preliminary subscales, using item analysis of the longer form. The
items in the test were multiple-choice items, with elements being

drawn from the Strong Vocational Interest Blank. The correct answer

 

for each item was that listed by Strong as having the highest preference
by persons of a given category (stereotype). Repeat reliabilities of
the stereotype scales and total scores ranged from . 37 to . 60. The
internal consistency, as measured by the Kuder-Richardson formula
#20, ranged from .16 to . 58.

Thirteen additional measures were obtained from one group
of 100 students to explore the relationships of the measures to the
scores on the several scales. A total of 266 college students served

as subjects in the various phases of the study.

Albert Zavala

Evidence was found to support the following conclusions.
The accuracy of a person's stereotype is related to his accuracy in
predicting the behavior of individuals (1 of . 20 in one test, and _r_ of . 29
in another). The accuracy of a person's stereotype has little relation
to his knowledge of psychology (3.. of . 14 with course grade in psychology).
The accuracy of a person's stereotype is related to his attitudes
towards methods of successful leadership (correlations range from
-. 24 to . 26).

As measured by the present instrument, stereotype accuracy
appeared to be a highly specific ability, for the correlations between
the subscales ranged from -. 19 to . 28. Whether this is an artifact of
the measuring instrument and whether instruments with greater
generality can be developed, are questions that can be resolved only

by further experimental inquiry.

A TEST OF STEREOTYPE ACCURACY

by

Albert Zavala

A THESIS

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

MASTER OF ARTS

Department of Ps ychology

1960

Este tesis se la dedico a mi madre con todo
mi cari'fio. Si no fuera por ella, esta obra no

conoce la lfiz del dia.

iii

ACKNOW LEDGEMENT

I would like to thank Dr. Henry Clay Smith,
my advisor, for the help and guidance without
which I would not have been able to carry this work

to its completion.

TABLE OF CONTENTS

INTRODUCTION ..........................................

HISTORY ..................................................

METHOD

DISCUSSION

SU MMARY AND C ONC LU SIONS

REFERENCES

APPENDICES

The College Man Scale

The Woman Stereotype Scale

The Sex Differences Scale

The Age Stereotype Scale

The Subjects

Preliminary Form of the College Man Scale
Experimental Form

Final Form

Additional Measures

RESULTS .............................................

Internal Consistency of the Subscales
Stability of the Subscales

Correlations between the Scales

Factorial Structure

Correlations with Trumbo and Cline Tests
Correlations with Course Grades
Correlations with Doré Leadership Test
Norms

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

10

10
ll
11
ll
12
l3
l3
14
15

18

l8
l8
19
20
20
22
22
24

26

3O

32

39

LIST OF TABLES

Table Page
1. Subjects .......................................... 12
2. Intercorrelations of the Scales ....................... l9

3. Factor Loadings for All 18 Variables on the Two

Factors Defined by the Stereotype Accuracy Scales ..... 21
4. Significant Correlations between the Stereotype Scales
and Other Variables ................................ 23

O

5. Norms for Industrial Psychology Students ............. 24

INTRODUCTION

The purpose of the present study was to develop a test of
stereotype accuracy, determine its homogeneity, internal consistency,
and stability, and correlate it with variables which might provide some
evidence of its validity. Stereotype accuracy, following the definition
of Cronbach (1955), is the ability to predict the typical responses of
a social group. Recent theoretical considerations and research results
indicate that a person's stereotype accuracy is of central importance
in contributing to the adequacy of his ”empathy, " "interpersonal
perception, " "clinical intuition, " and "sensitivity to others, ” to mention

a few of the concepts closely related to stereotype accuracy.

HISTORY

The history of the problem of stereotype accuracy starts
with the concept of empathy. The term was introduced by Lipps in
1907. Through the implicit or stated definitions of many writers
(Freud, as quoted in Allport, 1937; Baumgarten, 1931; Sullivan, 1940;
Fleiss, 1942; Reusch, 1948; Hutt and Miller, 1949; Fromm-Reichmann,
1950; Rogers, 1950), the definition of empathy has developed to the
form found in English and English (1958): "1 . . . attribution of the
feelings or attributes aroused by its surroundings (actual or depicted)
to a natural object or work of art . . . . 2. apprehension of the state
of mind of another person without feeling (as in sympathy) what the
other feels. "

Many psychologists have been concerned with theoretical
explanations of empathy (Luszki, 1951; Baxter, 1952; Halpern, 1954;
Norman and Leiding, 1956; Vernon and Stewart, 1956). Both clinical
psychologists (Escalona, 1945; Reusch and Prestwood, 1949;
McClelland, 1951; and Cowden, 1955), and industrial psychologists
(Kerr, 1951; Speroff, 1953, 1954; and Bell and Stolper, 1955) have
viewed empathic ability as being of considerable importance in the
areas of their interests. However, the analysis of experimental results
indicated that the measurement of empathy and the interpretation of

the measurements is peculiarly complicated.

Bender and Hastorf (1953) for example, found that a judge's
predictions were much more highly related to his own responses than
they were to the responses of the person whose behavior he was pre-
dicting. Thus, if it happened that the judge and the person being judged
were actually similar, this would improve the empathy scores of the
judge. In order to circumvent this problem, Bender and Hastorf
developed a "refined empathy" score by subtracting the judge's pro-
jection score from his prediction score. The projection score in this
case was the deviation of a predictor's own answers from his predictions
for his associate.

Norman and Leiding (1956) investigated the relationships
between the variables empathy and projection, and refined empathy

with regard to'individual and Mass Empathl tests. The Mass Empathy

 

 

test measured a person's ability to predict the typical self-responses
of a group of persons. This was called "raw empathy. " The number
of times his judgments for others' responses are the same as his own
responses are totalled and called "projection. " "Refined empathy" is
taken as "raw empathy” minus "projection. " Murstein (1957) gives a
detailed account of the weaknesses inherent in the measures used by
Norman and Leiding.

Cronbach (1955), in a more systematic and detailed analysis
of processes affecting scores on the understanding of others, concludes
that empathy scores are the sum of four components: Elevation,

Differential Elevation, Stereotype Accuracy, and Differential Accuracy.

He defines these components as:

Elevation (E). The elevation component has the form
(y. . - x. .) , y. . . is the average of J's predictions over all items and
all O's; it reflects his way of using the response scale. The elevation
component is increased by any difference between J's central tendency
of responding and the central tendency of the self-descriptions, for
all items and O's combined.

Differential Elevation (DE). Differential elevation reflects

 

how closely J's average prediction for O corresponds to 0'5 central
tendency of response, all items pooled and J's central tendency of
response held constant. That is, it reports J's ability to judge deviations
of the individual's elevation from the average. We may write DE in

this form:

DE.2:O'__ +o-__ - 20" r [11

2 . .
The variance cr_ expresses J's report of how much O's w111 differ
o-j
in elevation. This assumed dispersion in elevation will appear later

. . . 2 ,
as a component of the assumed s1m11ar1ty score. cr_ IS the true
x
o.

dispersion in elevation. The correlation r}? Y (to be symbolized
o' o-j

DEr) represents J's ability to judge which O's rate highest on the
elevation scale.

In some tests, central tendency of response (elevation)
reflects insignificant response sets. In other tests, elevation has an

important psychological meaning. Thus, if a high score on each item

indicates Morale, the correlation DEr shows how will J can judge
which O's say they have the highest Morale.

"Stereotype Accuracir (SA). Stereotype accuracy describes

 

J's ability to predict the norm for Os. It might be called 'accuracy
in predicting the generalized other. ' This score depends on J's
knowledge of the relative frequency or popularity of the possible
responses.

"We may write:

2 Z 2
SA, 2 0'— + a; - 2cr_ 03.: r_ 3? [2]
'1 y.ij .i y.ij .iy.ij.i

. . . . 2 .
Here each variance is computed over items. The variance cr_ IS
X

.i
the scatter of the actual means. SA represents ability both as to shape
and scatter. r- i (Stereotype Correlation, SA )represents accuracy
.. . r
. 13 . 1

in judging mean profile shape, i. e. the order of item difficulties.

"Differential Accuraczj (DA). Differential Accuracy reflects

 

ability to predict differences between Os on any item. This component
is averaged over items . . . this component . . . may be regarded as
combining variances with a correlation term. The correlation (DAr)
measures the ability to judge which Os have the highest scores on the
item, when the score is taken as a deviation from Os mean. There is
one such correlation for each item" (Cronbach, 1955).

Cronbach states that the failure to separate the components

of empathy makes interpretation ambiguous. Where errors of

measurement affect one element or component of a global score, they
influence the others also. Thus "significance tests are spurious and
correlations are artifactually raised or lowered" (Cronbach, in Taguiri
and Petrullo, 1958). Two of Cronbach's examples serve to clarify
these points:

One investigator hypothesized that the extent to which
teachers knew their pupils' opinions would correlate with their
effectiveness in establishing a good classroom atmosphere.

He asked pupils to rate their teacher on patience, friendliness,
etc. , and used the rating, averaged over pupils as a criterion.
Call this X. He also asked the teacher to predict what average
rating he would be given (Y), and used D, the absolute difference
between X and Y, as a measure to support the hypothesis that
superior teachers have greater insight--but this result occurred
simply because all teachers predicted a favorable Y for them-
selves. Predictions varied so little than in effect D differed
from Y only by a constant. If Y had varied substantially, there
would still have been some artifactual constraint on the correlation.

Linkages often arise in a disguise which makes them
difficult to trace. One example is the work of Hastorf and Bender.
They were concerned with "empathy, " "projection, " and "real
similarity, " as derived from self-perception of another person,
and their empathy score was influenced by the degree of real
similarity, and therefore constructed a more complicated dyadic
index called a "refined empathy score" instead of isolating inde-
pendent variables. This merely buried the artifacts deeper, as
Hastorf and Bender ultimately discovered.

Cronbach suggested that investigators reduce their measures to these
components to clarify such problems.

Following Cronbach's recommendations, Cline (1958), using
colored sound films of interviews, conducted a study of the variables
related to accuracy in interpersonal perception, where the scoring
system used "was specifically chosen to make possible the more

analytical studyof accuracy scores suggested by Cronbach. " Cline

found that the ability to judge responses of others is a general but
factorially complex process. He also found that Stereotype Accuracy
accounted for a large portion of the generality: "Perhaps the most
striking finding from the [Cline's] analysis is the importance of the
stereotype accuracy component. The tendency in previous investi—
gations has been however, to eliminate this component from accuracy
scores in order to have 'purer' measures . . . . This seems to be
an inappropriate emphasis . . . . It is really quite a surprising
finding that having an accurate stereotype contributes so much to
making accurate predictions about so diverse a group. It would there-
fore seem more appropriate to do research aimed at understanding

Stereotype Accuracy than to eliminate it from consideration" (Cline,

 

1958).

PROBLEM

The measure of stereotype accuracy suggested by Cronbach
and used by Cline is derived as a subscale from a film test that also
provides measures of other components affecting accuracy. Because
of the apparent importance of stereotype accuracy, the aim of the
present study was to develop a separate test of this component, and to
explore some of its possible correlates.

Speroff and Kerr (1951) made one of the earliest attempts
to develop such a measure. Expressing dissatisfaction with earlier
conceptions of empathy which neglected the ”prediction of behavior"
aspect, and focusing on the feelings and momentary experience of the
subject, they developed a test designed to measure more objective and
stable aspects of empathic ability. The Empathy Test (TET) consists
of three series of items which the subject is asked to rank in popularity
for various reference groups. Fourteen types of music are ranked by
the subject in order of preference of ”office workers"; 15 magazines
in order of circulation rates; 10 annoying habits in order of their
nuisance value for "persons over 40 years. " Reliabilities by test-
retest for this instrument range from . 67 to .83.

Theauthors present the results of correlation studies
supporting the validity of scores on their test: sociometric interpersonal

desirability (3 . 34), supervisory aptitude as measured by How Supervise?

 

(r's . 62 and . 32), campus offices held and number of semesters in

college (r's . 15), breadth of smile of degree candidates receiving
their diplomas (.5 . 64). It is the lone pencil and paper test in the
literature. But Thorndike (in Euros, 1959), says that ". . . the
relatively few studies [Bell, 1955; Siegel, 1954; and Van Zelst, 1953]
by persons not associated with the author have tended to yield pre-
dominately negative results. Unless the positive results reported

in the manual of the Empathy Test are verified in findings of other

 

workers, this test cannot be recommended as either a useful practical
device or a contribution to the description and understanding of the
individual. " It was the purpose of the present study, therefore, to
develop a test that would measure more general stereotypes than those

measured by Speroff and Kerr (1951), and that would be more valid.

10

METHOD

The manner in which each of the stereotype accuracy
scales was developed and the subjects employed are first considered.
This is followed by a discussion of the preliminary, experimental and
final forms of that test. The concluding sections discuss other
measures used in the study and correlated with the stereotype accuracy

scales.

The College Man Scale

The items for this subscale were drawn from the Strong

Vocational Interest Blank. E. K. Strong compared the interests of a

 

large number of persons in regard to occupations, amusements, and
activities. He tabulated the results in percentages for each activity,
amusement or interest. Each item in this scale consisted of four
interests. The correct answer for an item was that listed by Strong
as having the highest preference. For example, for college men in

general the percent who liked the following interests were:

Advertiser 53%
Politician 40%
Surgeon 46%
Aviator 65%
Poet 19%

Thus, several four-choice multiple-choice items can be generated

from these five elements. Of the four choices included in any item,

11
that choice (on the basis of Strong's figures), which has the highest

per cent as being liked, is the correct answer.

The Woman Stereotype Scale

The scale for the Woman Stereotype was developed in the
same manner as was the scale for the College Man Stereotype. The
only difference being that figures for the average woman were used as

the criteria for correct answers.

The Sex Differences Scale

The scale for the sex stereotype again was the same as
for the previous two scales. The criterion used in selecting which
element of an item is the correct answer, however, differs. In this
subscale, the subject picks that choice which is most popular among
women and the least popular with men. That is, he must choose that
answer which shows the greatest sex difference. Here again Strong's
figures are used as a reference. This scale is best defined as

measuring Differential Stereotype.

 

The Age Stereotype Scale

The Age Stereotype subscale was constructed similarly
to the College Man scale and the average Woman scale except that
age levels were used as choices instead of interest activities. In this

scale only one activity or interest was given in each item. The subject's

12
task is to answer by giving the age group for which he thinks that
activity is most popular. The possible age groups from which the
subject chooses are 15, 25, or 55 year-olds. The age categories

were also drawn from Strong.

The Subjects

All subjects in this study were students at Michigan State
University. The various groups used as subjects were tested during
their regular class periods. Table 1 gives a summary of the subjects
in each group, and the data obtained from them. The total number
of students in each group is given as well as the per cent males in

each group.

 

 

 

Table 1
Subjects
% .
Class Term N Data obtained
males

Industrial Psych. Fall 1959 89 84 Prelim. form
Industrial Psych. Winter 1959 100 85 Experimental form
Industrial Psych. Spring 1959 77 84 Final form
Industrial Psych. Spring 1959 77 84 Final form

(retest)

 

13

Preliminary Form of College Man Scale

As a pilot to the main study, the first form was prepared,
using 45 items. This form was administered to 80 students in the 1959
fall term Industrial Psychology course at Michigan State University.
The scores obtained ranged from 10 to 39 with a mean of 24. 95, a
standard deviation of 5. 88, and an internal consistency of _I_'_ . 66. The
test was readministered to 70 of the original 89 students three weeks
later and a test-retest correlation of . 64 obtained. The items in this
form were subjected to an item analysis, and the 30 best items were
retained. These items constituted the College Man Stereotype subscale

which was used in later forms in the study.

Experimental Forms

The preliminary form used in the pilot study was lengthened
by adding the subscales described earlier (Woman Stereotype, Sex
Stereotype, Age Stereotype). The Woman Stereotype subscale had
60 items as did the Sex Stereotype subscale. The 30 retained items
from item analysis of the preliminary form constituted the College
Man Stereotype subscale. Thus, the total number of items for the
experimental form was 198. This form was administered to 140
students in the elementary course of Industrial Psychology at Michigan
State University. Each scale was item analyzed separately to select
the 30 most discriminating items. The highest and lowest scoring

27% of the students (for each scale separately) constituted the dichotomy

14
used as the basis for the item analysis. The 30 items which showed
the greatest discrimination between the high and the low scoring groups
were retained, and a new scoring key was developed for the full scale
of 120 items. The answer sheets were then rescored with the new key.
The 100 students who took the experimental form of the
test also took several other tests during the term as part of an ongoing

research project of which this thesis is a part.

Final Form

The 120 items (30 in each of four scales) selected by
application of item analysis to the experimental form, constitutes the
final form (see Appendix B). The order for the correct answers within
each subscale was randomized to control for response set, where a
subject may have a tendency to favor choosing a certain element
number of a series of items. In the four-choice items each choice is
correct in approximately one-fourth of the items, and in the three-
choice items each choice is correct one-third of the time. The average
time taken to complete the entire scale is from 30 to 40 minutes. The
scales may be machine or hand scored, using IBM scoring keys.

The final form was administered to an Industrial Psychology
class, spring term. Seventy-seven students took the test at the

beginning of the course and six weeks later.

15

Additional Mea sur e 3

There were 13 additional measures available from the 100
winter term Industrial Psychology students. The relationships between
these measures and the scales were investigated to explore further the
nature of stereotype accuracy. The stereotype accuracy scores used
in the analysis were those obtained from the rescoring of the best 120

items of the experimental form. The measures obtained are:

1. Age.
2. Sex.
3. Grade in present Industrial Psychology course, consisting

of the sum of the first two examination scores.
4. American Council on Education Psychological Examination

for College Freshmen. The total score served as a measure of the

 

subject's general intelligence (Michigan State University, 1954-59).

5. Trumbo Test of the Ability To Predict Behavior. This test

 

contains six cases, each consists of a few sentences describing an
actual person. The subject predicts the behavior of this person by
answering 15 true-false statements about him. Actual behavior is used
as a scoring criterion. Corrected odd-even reliability of this 120-item
test was _r_ . 77 (Trumbo, 1955).

6. Suppression-repression scale. An unpublished scale
measuring the extent to which a person admits unpleasant or unfavorable

thoughts and behavior. Items are similar to the MMPI L (lie) scale.

l6
Validity and reliability for this scale have not been determined. This
was one of two criterion measures used in this study.

7. Cline Movie Prediction Test. This test is similar to the

 

Trumbo Prediction Test, except that the subject sees a short sound

 

color motion picture of a person being interviewed. The subject
answers several types of items predicting the behavior of the person.
The reliability of the test was . 71, computed from the Spearman-
Brown formula based on variances (Cline, 1958). Actual behavior

is used as the scoring criterion. This test was used as a criterion
measure for the present study.

8. Scientific Values Scale. An unpublished scale which
measures the extent to which a person believes in the value of science.
Internal consistency for this scale was .1; . 78, and the repeat reliability
after six weeks was also _I; . 78 (Hershey, 1958).

9. Doré Leadership Questionnaire. This is a test designed

 

to measure attitudes towards certain principles of leadership. It is
a four scale test with 30 items in each scale, items being of the forced-
choice type. Reliability was 3 . 85 (Doré, 1960).

10. Employee-Orientation. A subscale of the Doré leadership
test which measures the extent to which the subject thinks a leader
should consider the worker as an individual rather than as an instrument
to get out production. Reliability: _r_ . 78.

11. Delegation of Authority. Another subscale of the Doré

test which measures the extent to which the subject thinks the leader

17
should set up general conditions for workers and let them figure out
how the details of the work will be handled. Reliability: 3 . 86.
12. Differentiated Role. The third of the four subscales of

the Doré Leadership Questionnaire which measures the extent to which

 

the respondent believes that the leader should or should not perform
the same functions as the workers, plan the work, organize the work,
or help train the workers. Reliability for this scale was _r_ . 75.

13. Creates Teamwork. The fourth of Doré's leadership
test subscales. This measures the extent to which the subject thinks
a leader should work towards creating a spirit of teamwork within the
work group, has pride in the group, and feels satisfaction with the

accomplishments of the group. Reliability of this subscale was _I; . 68.

18

RESULTS

Below are summarized data dealing with the scales and
their relationships to each other. Following this are the results having
to do with their relationships to other measures. Then the results of
the validity tests are given. Finally, norms for college students are

reported.

Internal Consistency of the Subscales

The internal consistencies of the scales are reported below.
These were calculated by the Kuder-Richardson formula #20, using
data from the second administration of the 77 students in the spring

term Industrial Psychology class.

College Man Stereotype r . 51
Woman Stereotype _1; . 16
Sex Stereotype 3 . 42
Age Stereotype _r_ . 58
Total score . 3O

[*1

Stability of the Subscales

The test-retest reliabilities for the spring term class of
Industrial Psychology students, with six weeks between administration,

are given below. There were 77 students in this group.

term Industrial Psychology students (N = 100).

College Man Stereotype
Woman Stereotype

Sex Stereotype

Age Stereotype

Total score

IH

IH

IH

IH

IH

.60

.44

.46

.37

.56

Correlations between the Scales

l9

Intercorrelations of the scales were computed for the winter

summarize s thes e findings .

Table 2

The table below

Intercorrelations of the Scales

 

 

 

l 2 3
1. College Man Stereotype -
2. Woman Stereotype . ll -
3. Sex Stereotype —. 19 -. 08 -
4. Age Stereotype -, 01 . 00 . 28**

 

for all measures obtained from the winter term Industrial Psychology

students.

*‘i‘Significant at . 01 level (critical level for . 01 = . 26)

Appendix E reports the complete table of intercorrelations

20

Factorial Structure

In an effort to determine the simple structure underlying
the 18 measures obtained from the 100 winter term Industrial Psychology
students, a principal axis factor analysis, using a rotational procedure
suggested by Kiel and Wrigley (1960 APA Convention paper) was
performed. Quartimax solutions were used. The procedure involves
comparing solutions with increasing numbers of factors rotated until
stability is reached. The factor analysis and the correlations between
the 18 measures were computed on the MISTIC (Michigan State Iterative
Computer). Five factors were identified, two of which concern the
Stereotype Accuracy Test. Course grade, ACE scores, the Trumbo
test, College Man Stereotype, and Woman Stereotype were the major
variables identifying Factor I while Sex Stereotype, Age Stereotype and
the total scores for the Stereotype Accuracy Test were factors defining
Factor II. Table 3 reports the factor loadings for all variables on
these two factors. The complete table of loadings on all variables

for all factors can be found in Appendix C.

Correlations with Trumbo and Cline Tests

The Trumbo prediction test correlated . 20 with the total
score of the Stereotype Accuracy test. This was found to be significant
at the . 05 level. Subscale correlations with the Trumbo test were:
College Man Stereotype . 14, Woman Stereotype . 08, Sex Stereotype . 02,

and Age Stereotype . 13.

Table 3

Factor Loadings for All Eighteen Variables on the Two Factors

Defined by the Stereotype Accuracy Scales

 

 

 

Variable --I ----- Factor-—-I-I---
College Man Stereotype . 55 . 20
Woman Stereotype . 61 . 20
Sex Stereotype -. 37 . 58
Age Stereotype -. 09 . 81
Stereotype Accuracy, Total score . 32 . 89
Age -. 09 -. 31
Sex . 12 -. 04
Course Grade . 48 -. 02
ACE . 62 . 21
Trumbo . 33 . 23
Suppression-Repression . O7 . 00
Cline Movie . 04 . 45
Scientific Values -. 24 . 18
Differentiated Role . 25 -. 14
Employee Oriented -. 01 . O3
Delegation of Authority -. O8 -. 01
Creates Teamwork . 01 . 08

Leadership, Total score . 05 -. 01

 

22
The total score of the Stereotype Accuracy Test correlated
. 29 with the Cline Movie. Subscale correlations were found to be:
College Man Stereotype . 02, Woman Stereotype . 18, Sex Stereotype
. O7, and Age Stereotype . 32. Age Stereotype and the Cline Movie
correlated significantly beyond the . 01 level, as did the total score

for the Stereotype Accuracy Test.

Correlations with Course Grades

. A positive correlation was found between course grade in
the winter term Industrial Psychology group and the Woman Stereotype
subscale. Subscale correlations with course grade were: College Man
Stereotype . 07, Woman Stereotype . 20, Sex Stereotype . 07, Age

Stereotype . 02, and Total score . l4.

Correlations with Doré Leadership Test

Certain subscales of the Doré Leadership Test correlated
significantly with some of the subscales of the Stereotype Accuracy Test.
They are reported in this section.

Differentiated Role. The Differentiated Role subscale of the

 

Doré Leadership scale with the College Man Stereotype was found to be
. 20. This is significant at the . 05 level.

Delegation of Authority and Teamwork. These two subscales

 

 

correlated positively with the Sex Stereotype subscale. Delegation of

Authority correlated . 26; significant at the . 01 level. Creates Teamwork

correlated . 22;

23

significant at the . 05 level.

The table below summarizes all the findings of significant

correlations between scores on the Stereotype Accuracy Test and

other measures.

Table 4

Significant Correlations between the Stereotype

Scales and Other Variables

 

 

 

 

Related factors CMS WS SS AS Total
Age -. 21==
Sex -.25*
Course grade . 20*
ACE . 27*":< . 28**
Cline Movie . 32=z<>t< . 29>:<=:<
Trumbo . 20*
Differentiated Role , 20*
Delegation of Authority -. 24* , 26**
Creates Teamwork . 22*
Leadership Total . 21*

.05=.20 .01=.26

I»:

| *1

24

Norms

Norms for the final form of the test were developed, and are
summarized in Table 5. The norms reported here were developed from
data gathered from the spring term Industrial Psychology class at
IVIichigan State University. This class may be considered a represen-
tative sample inasmuch as there was only one such section taught that
term, and there is reason to believe that this group was typical of the

groups taking this course other terms. (See Table 1 also.)

Table 5
Norms for Industrial Psychology Students

N=77

 

 

First administration Second administration

 

Scale
Mean S. D. Mean S. D.
College Man Stereotype l6. l8 3. 6O 16. 35 4. 26
Woman Stereotype ll. 70 2. 81 11. 57 2. 65
Sex Stereotype 13. 16 3. 24 12. 74 3. 24
Age Stereotype 18. 39 2. 72 17. 79 3. 65

Total 59.42 5.97 58.43 9. 35

 

25

DISCUSSION

In spite of the weaknesses in measurement, some results
are consistent with findings from other studies and lend support to the
validity of the scales. Cline (1958) found that stereotype accuracy was
a major determinant of scores on his test; the present study found a
correlation of . 29 between Cline scores and stereotype accuracy scores.
Consistently, a correlation of . 20 was found between Trumbo's test
and stereotype accuracy scores. As in many tests of prediction
ability, the present study reports a correlation of . 28 between intelligence
and stereotype accuracy. Finally, Cowden's (1955) finding that males
are superior to females in ability to predict, is also supported by the
correlation of -. 25 between sex of the respondent and the College Man
Stereotype.

The findings of the present study, reviewed below, lend
little support to the still present possibility of finding a general, stable,
and valid measure of stereotype accuracy. They do, however, suggest
a new component of the problem as well as some further avenues to be
explored.

Stereotype scales had discouragingly low reliabilities.
Test-tetest reliabilities ranged from . 37 to . 60; the total score had
a test-retest reliability of . 56. Kuder-Richardson formula #20 internal

consistency reliabilities ranged from . 16 to . 58.

26
Scores on the subscales, which were expected to be at
least fairly closely related, were practically independent. Thus, as
measured by these scales, stereotype accuracy appears to be a highly
specific ability. Item analysis, by use of total scores, may lead to a
test of greater internal consistency. The significance of scores on a
test derived in this way is an open question.

Differential Stereotype Accuracy as a New Component
of Prediction Ability

This subscale measures the extent to which the respondent
can predict interests which show the greatest sex difference in terms of
preference. Inclusion of the scale has created theoretical and practical
problems from the beginning of the study. The subscale was included
primarily because of the convenient fact that Strong, in tabulating the
interests of people, reported those interests which differentiated men
and women most (Strong, 1943). The compatibility of this subscale
with Cronbach's definition of Stereotype Accuracy, therefore, was not
questioned until after most of the data had been gathered and analyzed.
A routine re-check of some of the literature on Stereotype Accuracy
revealed that the Sex Stereotype subscale was, in fact, not a measure
of Stereotype Accuracy. A closer examination of Cronbach's analysis
of the components of empathy was carried out in order to determine
what actually had been measured by this subscale. Here is where an

important finding was uncovered.

27

Cronbach’(l955) breaks empathy down into four components.
He used data from a study of a group of students at Cornell. His design
was two dimentional. None of the definitions for the components of
empathy in his analysis were suitable as models for what was measured
by the Sex Stereotype scale. This was simply because, if the measure
derived by the Sex Stereotype were applied to the Cronbach design, that
design would need to be extended by another dimension. This is no
criticism of Cronbach's design; he was not interested in any information
that could be gained by extending his design. Using Cronbach's analysis
as a paradigm, the Sex Stereotype subscale can be called Differential
Stereotype and described as follows:

Differential Stereotype (DS). This is a new concept that can

 

be considered a component of empathy. It is measured in this study by
the Sex Stereotype subscale. The Sex Stereotype subscale is not a

measure of Stereotype Accuracy, it is a measure of Differential

 

Stereotype. It reflects the ability of the judge to predict the differences

 

on any item (averaged over items) between groups of persons. This
definition differs from the Cronbach definition of Differential Accuracy
in that the Cronbach definition is concerned with differences between

individuals. Here the concern is with differences between groups.

 

This difference, averaged over items, is the factor which the judge must

predict accurately. We may write:

DS, 20' +0 -20' 0' r [3]

28
. . . 2 . .
The variance 18 computed over items. 0X" indicates
. i
the scatter of the actual means. This score depends on J's knowledge
of the differences in popularity or relative frequency of responses for
any item between groups. y" ij is J's description of the differences

between two groups on item i. 0' expresses J's report of how '

y"
. . ij
much one group of Os will differ from another group of Os in responding

to an item. It is the assumed similarity between groups. r n n 2 DS
x - r

and reflects J's ability to judge which group of Os have the . 13 . 1
highest scores on the item.

The concept may be clarified by considering the stereotypes
popularly attributed to Negroes and Jews. The same characteristics are
not attributed to both groups. Rather, differences are recognized in the
characteristics claimed of both groups, some of these are accurate and
others are not. Thus, it is conceivable that differences between the
responses of these two groups on any one item can be predicted. This
in fact is what was done in the case of SS, where males and females were
the dichotomy in question.

From our knowledge of analysis of variance, we know that
the component variances add up to one. Hence in the extended dimension
analysis of variance paradigm offered here, Differential Stereotype is a
component of empathy in the same sense as are the four analyzed by
Cronbach. In the present study no analysis of variance was actually
calculated.

The implications of this finding are several.

29

If the analysis offered in this study regarding Sex Stereotype
is correct, this would suggest that there are at least five components
of empathy, and perhaps more, depending on the complexity of the
analysis calculated.

The possibilities for combinations of groups whose differences
would be predicted are almost without limit. For example, the
differences between labor and management, Democrats and Republicans,
easterners and westerners, and men and women on a variety of issues,

as well as men and women in general.

30

SUMMARY AND C ON CLUSIONS

The purpose of this study was to develop an objective and
economical test that would measure Stereotype Accuracy, the ability
to predict the behavior of social groups.

Four 30—item scales were designed to measure the accuracy
of four stereotypes: College Man Stereotype, Woman Stereotype, Sex
Stereotype, and Age Stereotype. These were developed from four
preliminary subscales, using item analysis of the longer form. The
items in the test were multiple-choice items, with elements being

drawn from the StrongVocational Interest Blank. The correct answer

 

for each item was that listed by Strong as having the highest preference
by persons of a given category (stereotype). Repeat reliabilities of

the stereotype scales and total scores ranged from . 37 to . 60. The
internal consistency, as measured by the Kuder-Richardson formula #20,
ranged from . 16 to . 58.

Thirteen additional measures were obtained from one group
of 100 students to explore the relationships of the measures to the
scores on the several scales. A total of 266 college students served
as subjects in the various phases of the study.

Evidence was found to support the following conclusions.
The accuracy of a person's stereotype is related to his accuracy in
predicting the behavior of individuals (_r_ of . 20 in one test, and 1 of . 29

in another). The accuracy of a person's stereotype has little relation

31
to his knowledge of psychology (_1_' of . 14 with course grade in psychology).
The accuracy of a person's stereotype is related to his attitudes towards
methods of successful leadership (correlations ranged from -. 24 to . 26).

As measured by the present instrument, stereotype accuracy
appeared to be a highly specific ability, for the correlations between the
subscales ranged from —. 19 to . 28. Whether this is an artifact of the
measuring instrument and whether instruments with greater generality
can be developed, are questions that can be resolved only by further
experimental inquiry.

Several suggestions for future research emerge from this
study. First among them would be a program aimed at improving
reliabilities and validities of the instrument. For example, a correlational

study with the Speroff and Kerr Empatthe st would be of value.

 

Once the present test has been developed to the point that
it would be of service in applied as well as theoretical problems, a
series of tests, each measuring a separate component of empathy would
be highly desirable. For example a test measuring Differential
Stereotype, one measuring Differential Accuracy, and so on, would be
of value in re-investigating some of the earlier studies on empathy.
Perhaps in this way a better understanding of the process of empathy

can be attained.

32

REFERENCES

Allport, G. W. Personality, Ch. XIX, p. 532, New York: Holt &Co.,

 

1937.

Anikeef, Alexis M. Sociometric empathy. J. Social Psychology, 1957,

 

45, 283-287.
Baxter, Nathan. A study of the relationships between self-awareness

and social sensitivity. Dissertation Abstracts, 1952,

 

12, 719-720.

Baumgarten, F. Sympathy, understanding and empathy. Psychological

 

Abstracts, 1931, No. 241.
Bell, Graham B. , & Stolper, R. An attempt at validation of the

empathy test. J. Applied Psychology, 1955, 39, 442-443.

 

Bell, G. B., &Hall, H. E. Jr. The relationship between leadership

and empathy. J. Abnormal &Social Psychology, 1954,

 

49, 156-157.
Bender, I. E. &Hastorff, A. H. On measuring generalized empathic

ability (social sensitivity). J. Abnormal &Social Psychology,

 

1953, 48, 503-506.

Buros, O. K. (Ed.). Fifth mental measurements yearbook. Highland

 

Park, New Jersey: Gryphon Press.

Cline, V. B., & Richards, J. M. Variables related to accuracy in

 

interpersonal perception. First Annual Report, November,

 

1958, Contract No. NR 171-146, Group Psychology Branch,

Office of Naval Research, University of Utah.

33

Cowden, R. C. Empathy or projection? J. Clinical Psychology, 1955,

 

11, 188-190.
Cronbach, L. J. Processes affecting scores on understanding of others"

and "assumed similarity. " Pslrchological Bulletin, 1955

 

52, 177-193.
Crow, W. J., and Hammond, K. R. The generality of accuracy and

response sets in interpersonal perception. J. of Abnormal

 

and Social Psychology, 1957, 54, 384-390.

 

Crow, W. J. In defense of stereotypes, paper read at 1960 convention

 

of APA, Chicago, Ill.

Dailey, C. Unpublished Ph. D. dissertation, U. of M. , 1950.

Doré, R. L. The development and validation of forced choice scales
measuring attitudes toward leadership methods. Unpublished
master's thesis, M. S. U. , 1960.

Dymond, R. A preliminary investigation of the relation of insight and.

empathy. J. Consult.'Psychol., 1938, 12, 228-233.

 

Dymond, R. Personality and empathy. J. Consult. Psychol., 1950.

 

14, 343-350.

Escalona, S. Feeding disturbances in very young children. Amer.

 

J. Ortho., 1945,15, 76-80.

English, H. B. 8!. English, A. C. A comprehensive dictionary of psycho-

 

logical and psychoanalytical terms. New York: Longmans,

 

Green, 1958.

34

File, Q. W. The measurement of supervisory quality in industry.

J. Applied Psychol., 1945, 29, 323—337.

 

Fleiss, R. The metapsychology of the analyst. Psychiatric Quart. ,

 

1942, 211-227.
Foa, U. G. The contiguity principle in the structure of interpersonal

relationships. Human Relations, 1958, 11, 229-238.

 

Foa, U. G. Some correlates of the workers with the foreman. J.

Applied Psychol., 1960, 44, 6-11.

 

Force, R. C. , Empathy tests for selecting personnel to deal with

people in trouble. J. Soc. Ther., 1956, 2, 46-50.

 

 

Fromm-Reichmann, F. Principles of intensive psychotherapy. Chicago:
University of Chicago Press, 1950.

Gage, N. L. Explorations in the understanding of others. Educ.

 

Psychol. Measmt., 1953, 13, 14-26.

 

Gage, N. L., Leavitt, G. S. and Stone, G. C. The intermediary key in

the analysis of interpersonal perception. Psychol. Bull. ,

 

1956, 62, 411-461.
Guilford, J. P. Fundamental statistics in psychology and education.
New York: McGraw-Hill, 1956.

Halpern, H. M. Some factors involved in empathy. Dissertation

 

Abstracts, 1954, 14, 1957.

 

Hastroff, A. H. and Bender, I. E. A caution respecting the measure-

ment of empathic ability. J. Abnorm. Soc. Psychol. ,

 

1952, 47, 574-576.

35
Hastorff, A. H. 81 Bender, I. E. On measuring empathic ability

(Social sensitivity). J. Abnorm. & Soc. Psychol., 1953,

 

48, 503-506.

Heidbreder, E. Seven psychologies. New York: Appleton-Century,

 

1933, p. 101.

Hershey, G. L. College grades in relation to inventory measures of

 

personality. Unpublished master's thesis. Michigan State

 

University, 1958.
Hutt, M. &Miller, D. Social values and personality development, _._I_.

Social Issues, 1949, 5, No. 4.

 

Jarrard, L. E. Empathy: The concept and in-trial applications.

Personnel Psychology, 1956, 9, 157-158.

 

Johnson, R. J. Relationship of employee morale to ability to predict

responses. J. Applied Psychol., 1954; 38, 32-323.

 

Kerr, W. A. & Speroff, B. J. The empathy test. Chicago, Ill.:

 

Psychometric Affiliate s, 1951 .

Lipps, T. Das Wissen von Frenden Ichen. Psychologische

 

Untersuchungen, 1907, v. I, 694-722.
Luft, J. Differences in prediction based on hearing versus reading

verbatim clinical interviews. J. Consult. Psychol. ,

 

1951, 15, 115-119.
Luszki, M. Empathic ability and social perception. Ph. D. dissertation.

Dissertation Abstracts, University of Michigan, 1951.

 

36
McClelland, W. A. Preliminary test of role playing ability. J.

Consult. Psychol., 1951, 15, 102-108.

 

McMurry, R. N. Empathy: Management's greatest need. Advance

Management, 1953, 18(7); 6-11; 34.

 

IVIiller, F. G. 8: Remmers, H. H. Studies in industrial empathy, II.
Management's attitudes toward industrial supervision and

their estimates of labor attitudes. Personnel Psych. , 1950,

 

3, 33-40.
Murstein, B. I. Some comments on the measurement of projection

and empathy. J. Consult. Psych., 1957, 21, 81-82.

 

Nagle, B. F. Productivity, employee attitude and supervisor sensitivity.

Personnel Psychol., 1954, 7, 219-233.

 

Norman, R. D. & Leiding, W. C. The relationships and measures

of individual and mass empathy. J. Consult. Psychol. ,

 

1956, 20, 79-82.
Patton, W. M. Jr. Studies in industrial empathy III: A study of

supervisory empathy in the textile industry. J. Applied

 

Psych. , 1954, 38, 285-288.
Porter, L. W. & Kaufman, R. A. Relationships between a top-middle
management self-description scale and behavior in a group

situation. J. Applied Psych., 1959, 43: 345-348.

 

Remmers, H. H. A quantitative index of social-psychological empathy.

J. Orthopsychiat., 1950, 20, 16-165.

 

Remmers,

Rogers, C.
Reusch, J.
Reusch, . J.

37

L. J. &Remmers, H. H. Studies in industrial empathy 1:
Labor attitudes toward industrial supervision and their

estimate of management's attitudes. Personnel Psych. ,

 

1949, 2, 427-436.

S. Client Centered Therapy. Boston: Houghton-Mifflin,

 

1950, p. 29.
Experiments in psychotherapy, I Theoretical considerations.

J. Psychiat., 1948, 25, 137-169.

 

&Prestwood, A. Anxiety, its initiation, communication

and interpersonal management, Arch. Neurol. 81 Psychiat. ,

 

1948, 62, 527-550.,

Siegel, A. I. An experimental evaluation of the sensitivity of the

Speroff, B.
Speroff, B.
Speroff, B.
Stagner, R.

empathy test. J. Applied Psych., 1954, 38, 222-223.

 

J. Relationship between empathic ability and. supervisory

knowledge. J. Personnel Admin. in Industrial Relations,

 

1954, 1, 195-197.
J. Empathic ability and accident rate among steel workers.

Personnel Psychol., 1953, 6, 297-300.

 

J. Empathy and role reversal as factors in industrial

harmony. J. Soc. Psychol., 1953, 37, 117-120.

 

Psychological aspects of industrial conflict. Personnel

 

Psychol., 1948, 1, 131-143.

Stone, G. C., Gage, N. L. & Leavitt, G. S. Two kinds of accuracy

in predicting another's responses. J. Soc. Psychol., 1957,

 

45, 245—254.

38

Strong, E. K. Vocational interests of men and women. Stanford:

 

Stanford University Press, 1943.

 

Sullivan, H. S. Conceptions of modern psychiatg. Washington, D.C.:
W. A. White Foundation, 1940, p. 6.

Taft, R. The ability to judge people. Psych. Bulletin, 1955, 52, 1-22.

 

Taguiri, R. & Petrullo, L. Person perception and interpersonal

 

behavior. Stanford: Stanford Univ. Press, 1958.

Thorndike, R. L. Personnel selection. New York: Wiley, 1949.

 

Trumbo, D. A. The development and analysis of a test of the ability to

predict behavior. Unpublished master's thesis. Michigan

 

State University, 1955.

Van Zelst, R. H. Empathy test scores of union leaders. J. Applied

 

Psych., 1953, 36, 293—295.

Van Zelst, R. H. Validation evidence on the empathy test. ESE
Magi, 1953, 13, 474-477.

Vernon, G. M. 81 Stewart, R. L. Empathy in the dating situation.

Amer. Soc. Rev., 1957, 22(1), 48-52.

 

Wrigley, C. F. , &Neuhaus, J. The quartimax method: An approach

to the orthogonal simple structure. Brit. J. Stat. Psychol.

 

Nov. , 1954.

APPENDIX A

39

H. C. Smith
January. 1960

STEREOTYPE ACL'URACY

Much of our ability to Judge a specific person depends on our under-
standing of the general type to which the person belongs: Our understanding
of a particular college man depends on our understanding of college men in
general; our understanding of a particular woman depends on our understand-
ing of women in general: our understanding of the differences between a
particular man and woman depends on our mderstanding of the differences
between men and women in general; our understanding of the differences
between a particular young man and particular old man depemis on our
understanding of the differences between young and old men in general;
etc.. The four tests below measure your general understanding of the
interests of women and college men. of the different interests of men and
women. and of the different interests of young and old people.

Collegg Mg Stereotng

Undergraduate men in a Midwestern university were asked whether they liked
or disliked each of a large number of different occupations, school subjects.
amusements. and kinds of people . Mark on the separate answer sheet which
one of the four in the following groups you think was liked by the most
QtudODts e

25.

(1) artist (2) army officer (3) author of a novel (11) pharmacist.

(l) surgeon (2) mechanical engineer (3) statistician (it) interpreter.

(l) astronomer (2) athletic director (3) sculptor (a) toolmaker.

(1) watchmaker (2) rancher (3) landscape gardener (h) mining superintendent.
(l) undertaker (2) cashier in bank (3) music teacher (11) dentist.

(1) magazine writer (2) chemist (3) retailer (lb) secret service man.

(1) ship officer (2) carpenter ( 3) laboratory technician (ll) railway conductor.
(1) printer (2) scientific research worker (3) judge (ll) orchestra conductor.
(1) laboratory technician (2) politician (3) private secretary (is) toolmaker.
(1) photo engraver (2) jeweler (3) ship officer (1+) explorer.

(1) poet (2) printer (3) scientific research worker (a) governor of a state.
(1) secret service men (2) farmer (3) auto racer (h) inventor.

(1) army officer (2) electrical engineer (3) bookkeeper (it) aviator.

21) history (2) calculus (3) algebra (h) chemistry.

1) nature study (2) physics (3) economics (ll) military drill.

(1) geology (2) zoology (3) geometry (h) philosophy.

(1) English composition (2) mechanical drawing (3) mathematics (a) psychology
(1) art (2) geography (3) physics (1*) zoology. .
(1) military drill (2) shop work (3) public speaking (1+) calculus.

(1) civics (2) geology (3) dramatics (4) languages, ancient.

(1) ”New Republic" (2) "System” (3) "Time" (4) "Popular Mechanics!

(l) musical comedy (2) symphony concerts (3) vaudeville (l4) fortune tellers.
(1) solving mechanical puzzles (2) full dress affairs (3) auctions

(1+) performing sleight-of-hand tricks.

(1) repairing electrical wiring (2) cabinet making (3) interviewing clients
(1+) climbing along edge of precipice.

(l) decorating a room with flowers (2) contributing to charity (3) adjusting
a carburetor (l1) adjusting difficulties of others.

-2-

(l) continually changing activities (2) displaying merchandise in a store

( 3) expressing judgments publicly regardless of criticism (’4) looking at

a collection of antigue furniture.

(1) repairing a clock (2) looking at a collection of rare laces (3) pursuing
bandits in sheriff's posse (h) entertaining others.

(1) foreigners (2) fashionably dressed people (3) independents in politics
(h) energetic people.

(1) conservative people (2) optimists (3) people with gold teeth (a) talk—
Cti'e people.

(1) peeple who chew gum (2) people who talk very slowly (3) people who are
natural leaders (1+) thrifty people .

W

E. 1:. Strong3 asked several thousand women whether they liked the activities
involved in a variety of occupations and amusements. and certain types of people.
He then calculated the percent of wanen who liked each. Mark on the separate
answer sheet the one of the four interests in each of the groups below that you
think was liked by the met women.

31.
32.
33-
34.
35.
36.
37.
38.
39.
110.
1+1.
1&2.
1&3.
101‘.
1+5.
46.
47.
1&8.
“9.
50.
51.
52.
53.
5“.

(1) Architect (2) advertiser (3) accomtant (1+) movie actress.

(1) Stage actress (2% artist (3) author of children's books (a) aviatrix.

(1) artist's model 2) athletic director (3) author of novel (4) bank teller.

(1) buyer of merchandise (2) biologist (3) bacteriologist (1b) bookkeeper.

(l) costume designer (2) cook (3) confectioner (h) cashier.

(l) dietitian (2) Dean of Women (3) dramatist (ll) florist.

(1) cartoonist (2) caterer (3) chendst (1}) foreign correspondent.

(1) college professor (2) employment manager (3) editor (h) interior decorator

(1) interpreter (2) inventor (3) landscape gardner (ll) judge.

(1) magazine writer (2) librarian (3) laboratory technician (h) illustrator.

(l) opera singer (2) mic composer (3) musician (1+) hostess.

(1) physician (2) minister (3) nurse (1;) naturalist. ‘

(l) playground director (2) poet (3) office manager (1+) milliner.

(1) professional dancer (2) private secretary (3) psychiatrist (h) publisher.

(1) psychologist (2) social worker (3) surgeon (It) eductional director.

(1) musician (2) wife (3) interior decorator (1+) psychologist.

(l) vocational counselor (2) tea room proprietor (3) sculptress (it) poet.

(1) dancing (2) camping (3) playing a musical instrument (U) animal 2008.

(1) riding horses (2) golf (3) swiming (1+) bridge.

(1) poetry (2) conventions (3) driving an automobile (it) afternoon teas.

(1) picnics (2) tennis ( 3) forum]. affairs (4) fortune tellers.

(l) dancing (2) swimdng (3) plays (it) picnics.

(1) art galleries (2) amusement parks (3) conventions (1+) attending lectures.

(l) poker (2) conventions (3) animal zoos (4) art galleries.

(1) museums (2) golf (3) detective stories (4) fortune tellers.

(l) riding horses (2) musical comedy (3) bridge (a) observing birds.

(1) symphony concerts (2) bridge (3) poker (1+) afternoon teas.

(l) fortune tellers (2) animal zoos (3) attending lecturers (11») movies.

(211‘; "House and Garden" magazine (2) romantic stories (3) movie magazines
poetry.

(1) "National Geographic Magazine” (2) "Vanity Fair" (3) ”True Story“ (10pm.

(1) "Reader's Digest" (2) "New Republic“ (3) "House and Garden“ (4) ”True Story"

(1) Energetic people (2) conservative people (3) people who assume leadership

(ll) independents in politics .

(1) people who take life seriously (2) foreigners ( 3) optimists (h) athletic

women.

-3-

615. (1) foreigners (2) religious people (3) methodical people (1.) cautious people.

65. (l) emotional people (2) people who are natural leaders (3) self-conscious
people (h) fashionably dressed people.

66. (1) people who are unconventional (2) men who drink (3) people who have made
fortunes in business (lb) thrifty people.

67. (1) people who take chances on situations of doubtful outcome (2) athletic
women (3) thrifty people (4) maple who have done you favors.

68. (1) religious people (2) people who are unconventional (3) people who take
life seriously (it) sick people.

69. (l)negroes (2) sick people (3) methodical people (‘5) fashionably dressed people.

70. (1) very old people (2) absent-minded people (3) cautious people (11-) people
who tell you their troubles .

71. (1) men who drink (2) conservative people (3) nervous people (4) people
with physical disabilities.

72. (l) calculus (2) algebra (3) physics (4) shorthand.

73. (1) art (2) bible study (3) bookkeeping (1.) Journalism.

71}. (l) chemistry (2) civics ( 3) English composition (15) political science.

75. (l) geography (2) geoloy (3) geometry (’4) domestic science.

76. (1) sociology (2) history (3) public speaking (1.) physical training.

77. (1) modern languages (2) dramatics (3) ancient languages ('4) physics.

78. (1) pmiol (2) literature (3) art (1») philosophy.

79. (1) music ((33 botany (3) economics (13») sociology.

80. (1) physics (2) geology (3) mechanical drawing (1.) nature study.

81. (1) philosophy (2) pennanship (3) public speaking (h) typewriting.

82. (1) statistics (2) political science (3) ptvsical training (a) shorthand.

83. (l) chemistry (2) physiology (3) ecology (lb) chemistry.

84. (1) arithmetic (2) psychology ( 3) economics (14) English composition.

85. (l) spelling (2) Journalism (3) teacher training (a) typewriting.

86. (1) decorating a room with flowers (2) attending church (3) preparing dinner
for guests (4) giving "first-aid" assistance.

87. (1) reading editorial columns (2) W at an auction sale (3) saving
money (1+) taking responsibility.

88. (1) raising money for a charity (2) entertaining others (3) meeting new
situations (it) contributing to charities.

89. (1) trying new cooking recipes (2) looking at shop windows (3) meeting and
directing people (lb) writing a personal letter.

90. (1) regular hours of work (2) writing reports (3) making a speech (4) cooldne.

Se; Stereom

E. K. Strong compared the interests of a large number of women and men. About
the same nunber of women and men expressed a liking for most occupations, amuse-
ments. and activities. But for some interests more women expressed a liking than
men. Mark on the separate answer sheet the interest in each of the groups below
that showed the greatest sex difference. Mark. in other words. the one interest
that you think the M34 women liked and the fewest men.

91. (1) artist (2) accountant (3) advertiser (it) architect.
92. (l) bacteriologist (2) athletic director (3) author of novel (h) bank teller.
93. (l) biologist (2) bookkeeper (3) cartoonist (h) cashier.
91+. (1) bwer of merchandise (2) cook (3) chemist (it) confectioner.
95. (1) costune designer (2) Dean of Women (3) dietitian (it) florist.
. (1) employment manager (2) factory manager (3) factory worker (14) tanner
97. (1) educational director (2) foreign correspondent (3) editor (4) hotel manager.
98. (1) interior decorator (2) housekeeper (3) illustrator (h) dress maker.
99. (l) lawyer (2) Judge (3) into ter (lb) laboratory technician.
100. (1) missionary (2) minister (3 milliner (h) landscape gardener.

101..
102.
103.
101+.
105.
106.
107 .

108.
109.
110.
111.
112 .
1.13.
11“.
115.

116.
117.

-14...

(l) musician (2) opera singer (3) music composer (it) museum director.

(1) playground director (2) physician (3) pharmacist (h) office manager.

(1) proof reader (2) post (3) radio singer (it) professional dancer.

(1) social worker (2) probation officer (3) psychologist (1+) psychiatrist.

(l) reporter (2) registrar (3) scenario writer (1+) private secretary.

(1) social worker (2) telephone operator (3) art teacher (it) statistician.

(l) vocational counselor (2) high school teacher (3) stenographer

(u) m tanner.

(l) typist (2) stenographer (3) telephone operator (h) comnercial teacher.
(1) taking long walks (2) dancing (3) winning (it) camping.

(1) golf (2) tennis (3) riding horses (14») driving an automobile.

(1) bridge (2) poker (3) playing a musical instrlment (h) afternoon toes.
(1) animal zoos (2) amusement parks (3) Picnics (u) conventions.

(1) plays (2) movies (3) fortune tellers (h) formal affairs.

(1) museums (2) attending lectures (3) art galleries (it) observing birds.
(1) musical comedy (2) ranantic stories (3) detective stories (1+) movie
magazines.

(1) symphony concerts (2) plays (3) movies (ll) attending lectures.

(1) movie magazines (2) detective stories (3) romantic stories (4) poetry.
(1) "ladies Horne Journal” (2) "National Geographic Magazine a

(3) "Atlantic Monthly" ('4) "House and Garden" magazine.

(1) operating machinery (2) giving "first-aid" assistance (3) attending
church (14») making a speech.

(1) raising flowers and vegetables (2) being the first to wear the very latest
fashions (3) repairing electric wiring (it) interviewing clients.

(1) argunents (2) making a speech (3) teaching adults (b) decorating a

room With HOW”.

(I) organizing a play (2) preparing dinner for guests (3) meeting and
directing people (it) taking responsibility.

(1) teaching children (2) teaching adults (3) discussions of economic affairs
(1+) discussions of politics.

(1) writing reports (2) entertaining others (3) acting as yell-leader

(1+) adjusting difficulties of others.

(1) writing personal letters (2) buying at an auction sale ( 3) trying new
cooking recipes (1+) doing research work.

(1) saving money (2) regular hours of work (3) looking at shop window8

(it) preparing dinner for guests.

(1) contributing to charities (2) saving money (3) taking responsibility

(h) meeting new situations.

(1) expressing judgments publicly. regardless of criticism (2) raising

money for charity (3) being head of a civic improvmnent program
(1+) discussions of politics.

(1) looking at a collection of rare laces (2) organizing a play
taining others ('4) contributing to charities.

(l) progressive people (2) conservative people (3) people who borrow things
(it) energetic people .

(1) absent-minded people (2) very self-confident people (3) pessimists

(1+) people who are natural leaders.

(1) people who assume leadership (2) very intellectual people (3) emotional
people (ll) thrifty people.

(1) people who have done you favors (2) people who take life seriously

(3) witty people (1.) negrces.

(l) cautious people (2) sick people ( 3) people with physical disabilities
(h) people who always agree with you.

(1) people who tell you their troubles (2) people who talk very loudly

(3) people who talk about themselves (h) methodical people.

(3) enter-

78.
79,

-h- as
(l) contributing to charities (2) organizing a play (3) entertain-
1 others (A) looking at a collection of rare laces.
(l? People who assume leadership (2) very intellectual people

(3) emotional people (a) thrifty people.
(1) people who tell you their troubles (2)'people who talk very

‘ loudly (3) people who talk about themselves.

as,

87.

(l) independents in politics (2) carelessly dressed people

(3) "Mannishfl women (h) fashionably dressed people.

(1) nervous people (25 people who chew gum.(3) men who drink
(h; women who smoke. "

(l ‘athlotic women (2) women cleverer than you are (3) cautious
people (E7 optimists. "” " “ ‘ " " ' '
11)'efiergetic people (2) religious people (3) very intellectual

' people (h) abscntaminded people.

(1) thrifty people (2) self-conscious people (3) people who are

natural leaders (h) foreigners.

(l) displaying merchandise in a store (2) being left'to yourself
(3) continually chan,ing activities (k) saving money

(1) making a speech 2) reading editorial columns (35 opening a

° conversation with a stranger (k) taking responsibility.

88
89.

90.

(l) president of a society (2) secretary of a society (3) treas-
urer of a society (k) member of a socie . '

(1) Chairman Entertainment Committee (2 Chairman, Educational
Committee (33 Chairman, Membership Committee (H) Chairman, Ar-
rangement Committee. " "

(1) draftsman (2) factory worker (3) chemist (k) Buyer of merchandis

_Age Stereotypes

E. K. Strong studied the changing interests of men from youth to old

age.

Cn'the”whole, he’found that interhsts changed relatively little

'with age. He did find, however, that lS-ycsr olds, ZS-year olds, and
55-year olds had.some distinctive interests. Fbr each of the inter-

ests below mark:

91;
92;
93:
9k;
96:
97;
98;
99:
100;
101;
102;
103;
10h;
105;
106;
107;
108.

._- O

(1) If you think it was liked most by IS—yenr olds;
C2) If you think it was liked most by ZS-year olds;
(3) If you think it was liked most by 55-year olds.

Civil engineer

Auto racer

machinist

Auctioneer

Poet

Social worker
Statistician
Politician

Florist

Auto repairman

Life insurance salesman
Hunting °

Thomas A. Edison
Popular Mechanicgf
Progressive people
Saving money ' '
Energetic people

People who are natural leaders

166.
16?.
168.
169.
170.
171.
172.
173.
174.
175.
176.
177.
178.
179.
180.
181.
182.
183.

185.
186.
187.
188.
189..
190 .
191.
192.
193.
194.
195.
196.
197 .
198.

Hunting

Athletic men

Fishing

Thomas A. Edison

mg» mghmics

Progressive people

Operating new machine

Saving money

Energetic people

People who are natural leaders
Handling horses

Travel movies

Museums

Educational movies

Musical comecw

Sporting page:

Discussing my ideals with others
Thrifty people

People who talk very loudly
Decorating a room with flowers
Teaching adults

Economics

President of a society or club
Chairman. entertainment committee
mailman. publicity cocnittee
looking at a collection of rare laces
szdeville

Egg Remblic

Men who chew tobacco
Electrical engineer

Floor walker

Specialty salesman

Life insurance salesman

1-18-60 81

109;
110;
111:
112.
113.
1111,.
115.
116;
117.
118;
119.
120.

2/16/60

Handling horses

Travel'movies

Museums

Musical comedy

Discussing my ideals with others
Thrifty‘people

People who talk very loudly
Decorating a room.with flowers
Economics '
President"of a society or club
Vaudeville'

Flocrwalker

EZ

APPENDIX C

52

Rotated Factor Loadings for SA and Other Measures

53

 

 

Variable I II III IV V
Age . 036 -. 086 . 314 -.i§4_ -. 240
Sex .136 .124 . 045 .823 -. 148
Course Grade . 434 . 183 . 024 . 066 .187
ACE .122 ._6_1_5_ .212 .108 .306
Trumbo Prediction . 330 .22 . 229 -. 089 . 308
S-I . 024 . 074 . 002 -._14_1_ -. 080
Cline Movie . 015 . 040 . 453 . 133 -._4_7_4_
Sci. Val. .120 -. 236 .180 -. 092.61
Differentiated Role . 352 . 250 . 138 -. 247 -._Ԥ_0_
Employee Oriented .810 - 010 . 027 -. 008 . 040
Delegationof Authority ._8_2_f1_ -. 076 . 013 . 206 . 091
Creates Teamwork .221 . 008 . 082 -. 024 . 018
Total (Doré Test) .183 . 053 . 010 -. 004 -. 091
College Man Stereotype . 112 ._5_42 . 199 -. 458 -. 072
Woman Stereotype . 010 .621 . 204 . 220 . 004
Sex Stereotype . 279 -. 373 57—6 . 240 . 065
Age Stereotype . 055 -. 095 8_1_3_ -. 049 -. 117
Total (Stereotype Accuracy) . 037 . 319 892 -. 064 —. O69

 

APPENDIX D

54

55

Notations

Notations in the text follow those of Cronbach, except for

notation number 8, which is an adaptation from Cronbach's and makes

up the formula used to identify the new concept introduced in this paper.

x0, Self-description of Other 0 on item i
1 — —

yoij Judge J_'_s description of _o on_i_

3: Average of x , over all items
0 01

x i Average over Others
x. . Grand mean over Others and items. _y is defined similarly.

x , = x , - 3? - 3? , +3? the score x . transformed as a
C O O C 01

deviation from both item mean and Other mean; yoij similarly.

(y .. - X .) is error in prediction
or] 01

y Judge _j_'_s description of the differences between two

'1‘.le

groups on item i.

APPENDIX E

56

57

3mm “ammo >Ho>o mo “50: a: “com Hmefioop m oESmw<

3 . 2 1.52 S. 8o 2:9 H8320:

om . 3 3R: mo. H8 239 183.5,...

 

 

 

 

. co nv wv wv 00 Ma No: vo no mo am no on wN wfi Ho wan .wH

u mm 00 Ho- mo- Ho: we: mo: no: mo Nm wo pd Ma No: 00 van .hd

1 mo- oHu Hm NN 0N pH mo: wH no wan No 00 no 0H HM: .oH

1 Ha vol Ma No 00 mo: mo: mg 00: mo hm ON Na o0: .mH

u mo: #0: “Hum: No: ON no: No «L. a; Na no mm: Ho 4;

u w» Nw Pm av 00 $0 No: hm MM mm wfi mo .mH

. cw we 0H Ma No Ho NH NH mm 0H Ho .NH

. mo NH OH do «H: mm AH mm ad mo- .HH

. MA oo $0 No: mm NH MN mo mo- .oH

u Md: we ma oo v0 mm 00 mg .m

u no: Han mo mo Ha NH: mo- .m

u can mo No mo- mo co: .>

I oo «0 we: mwu Hm .o

u 0N om we: NH: .m

u MN ma mm- .w

n ma we .m

.. wmu .N

n .H

.L m m1 n . m . mum WV... . a .m S
wmsmmwaéamemLm. v e m a

ac e r Lr r ea 1m Pm HR .L .m R u E .a x e
tc no a OmW OMW o r e p.u m r .io c .1 . r AC r e as
SA A S WS CS D CT DA E0 DR S C S T A G S A

00H M Z .muaoodum >mo~ozo>mna 8.38 young? no“ mondmmoz :< Go mcoflmaounoonouca mo 3an 303800

.‘

v."

 

HICHIGQN STRTE UNIV LIBRQRIES

03596452