The Journal of Social Psychology, 1979, 109, 253-261.

A MEASURE OF COOPERATIVE, COMPETITIVE, AND

INDIVIDUALISTIC ATTITUDES*
University of Minnesota

Davip W. JOHNSON AND ARDyTH A. NOREM-HEBEISEN

SUMMARY

Six studies involving over 6000 students from kindergarten through

college were sequenced over a period of years to develop threerelatively
independentself-report scales with substantial internal reliability to mea-
sure attitudes toward cooperative, competitive, and individualistic inter-
dependence between oneself and others in educational settings. The pur-
pose of the scale development was to provide a research tool for social
scientists interested in social interdependence. In addition, evidence consis-
tent with a multidimensional view of social interdependence was gathered,
and some additional evidence indicating a change in conceptions of the
three types of social interdependence in junior high school was found.

A. INTRODUCTION
There are several hundred studies comparing the relative effects of
cooperative, competitive, and individualistic situations (5, 6). These three
types of social interdependence may be defined asfollows (3, 5): Coopera-
tive interdependence involves a positive correlation among goalattain-
mentsof the people involved; competitive interdependence involves a nega-
tive correlation among the goal attainments of the people involved; and
when the goal attainments of people are independent, an individualistic
situation exists. Cooperative, competitive, and individualistic conditions
have been found to be related to performance in achievementsituations,
self-esteem, achievement motivation, attitudes toward other students, at-
titudes toward school and subject areas, attitudes toward teachers and
other school personnel, locus of control, perspective-taking ability, psycho-
logical health, and many other important aspects of cognitive and social
development. Despite the proliferation of studies of cooperative, competi-

* Received in the Editorial Office, Provincetown, Massachusetts, on July 5, 1978. Copy-

right, 1979, by The Journal Press.
253

Copyright © 2001. All Rights Reseved.

254 JOURNAL OF SOCIAL PSYCHOLOGY

tive, and individualistic situations, there is an absence of attitude scales

that measure positive or negative dispositions toward the three types of
social interdependence. This article seeks to document the developmentof
self-report measures of attitudes toward cooperative, competitive, and in-
dividualistic social interdependence for use in instructional settings.
The development of scales measuring attitudes toward social interdepen-
dence provides an opportunity to examine their theoretical relationship.
Deutsch’s (2) original theory assumed that cooperation and competition
were opposite ends of a single dimension. Individualistic situations were
added in later research with the implication that they were a midpoint
between cooperative and competitive situations. Despite the hundreds of
studies conducted on cooperative, competitive, and individualistic situa-
tions, there is little evidence for clarifying whether or not the three consti-
tute a continuum along a single dimension. Alternatively, the three types of
social interdependence may reflect two or three separate dimensions.

B. METHOD
1. Samples
The samples were, respectively, the entire student population (over 6000
students from grades 1 through 12) of a suburban school district, 143
12th-graders, 241 7th-graders, 153 12th-graders, 270 7th-graders and 152
college undergraduates. All samples were from a metropolitan area in the
Midwest. With the exception of the first sample, each population was a
nonprobability sample involving voluntary participation of all students
within collaborating teachers’ classes. Anonymity of each student was
assured.
2. Scale Development
The original development of the scales was part of the development of a
larger instrumentto be used for evaluation of affective outcomes of schools.
As part of the Minnesota School Affect Assessment (MSAA), attitude items
containing a four-point false-true range were generated to measure such
factors as cooperation, competition, and individualistic efforts, importance
of school, attitudes toward the teacher, and locus of control. The items
were administered to all students in grades 1 through 12 in a large subur-
ban school district in the Midwest (4, 7). The results were subjected to
principal component factor analyses with varimax rotations; among the
scales resulting were measures of cooperative, competitive, and individ-
ualistic interdependence with three items perscale.
 D. W. JOHNSON AND A. A. NOREM-HEBEISEN 255

Subsequently, five studies were sequenced over three years to develop

the attitude scales further. Characterized by recursive cycles of concep-
tualization, scale development, administration to Ss, and data analysis, the
studies were implemented to produce a theoretically sound, internally
reliable, self-report measure of attitudes toward social interdependence.
The measures of social interdependence changed from Study 1 to Study 6
in several ways: there were more items per scale, the items were more
descriptive, there were fewer scale-in-an-item statements (such as, “I like to
compete”), and the content of the scales became broadened to include the
advantages and value of each type of social interdependence as well as the
liking of them. Reading vocabulary level of the items was limited to the
5th-grade level. In the second study a 12-item array with content congruent
with the three types of social interdependence was administered to 143 high
school seniors. In the third study the item array was broadenedto 26 items
and administered to 241 7th-grade students. In the fourth study the item
array was broadened to 32 items and administered to 153 high school
seniors. In the fifth study the 32 item format was administered to 270
7th-grade students and, finally, the 22 items selected to be in the three
scales were administered to 152 college undergraduates. The response
format for all forms of the instrument in the last five studies was a
false-true scale with a seven-point range.
In the second through the sixth studies five issues were addressed: the
congruence of the factor structure with the theoretical constructs, the
stability of the factor structure across method of factor analysis, the stabil-
ity of the factor structure across age and sex subgroups, the internal
reliability of the item clusters designated for the construct, and the inde-
pendence amongthe three scales. As the item array used in the analysis
was expanded during the second, third, and fourth studies, responses for
each of the populations were analyzed separately. Student ratings to each
item array were subjected to Alpha factor analysis with varimax rotation.
Those factors with at least three or more rotated factor loadings = .5 and
eigen values = 1 were identified. Items with loadings = .5 were initial
designative items and identified with logical content of each factor. High
loading items for each factor comprised the initial item cluster, additional
items for each cluster were identified by use of iterative procedures to
maximize internal reliability for the cluster. When all high loading items
that maximized reliability were identified, they were forced into a three
factor structure to check for the robustness of the three component model.
Alpha factor analysis and Rao’s canonical factor analysis were compared in

Copyright © 2001. All Rights Reseved.

256 JOURNAL OF SOCIAL PSYCHOLOGY

one group of Ss, the 153 high school seniors, to determine generalizability
of the factor analysis across factor methodology. Other analyses made to
test for generalizability included separate Alpha factor analyses across sex
and age groups, Cronbach's alpha measure of internal reliability, and
correlations among the three scales.

C. RESULTS
Given below, in the following order, are the descriptions of the results of
the factor analyses, the examination of factor structures separately for
males and females, a test of factor generalizability, a final factor analysis
on the items designated for use in the refined scales, Cronbach alpha
indices, and interscale correlations.
To heighten the sensitivity of the analysis to dimensions within the item
set, initial factor analyses on the responses of each sample were conducted
without specifying the number of factors to be generated; the only criterion
was that all factors have eigen values= 1. Among the two 12th-grade
samples, analysis conducted under these specifications produced three fac-
tors with three or more items loading= .5 on each factor. High loading
items for these factors were used to designate the factors’ contents; the
logical components were consistently cooperative, competitive, and indi-
vidualistic interdependence between oneself and others. By contrast, factor
analyses without specification of the number of factors in the 7th-grade
samples produced up to four or five factors with three or four items with
factor loadings = .5 on each factor. Whenthe logical contents of these item
sets were examined, it appeared that logical subsets of cooperative, compet-
itive, and individualistic social interdependence had emerged: liking to
cooperate and valuing cooperative learning, liking to compete and valuing
competitive learning, and liking to study alone and valuing individualistic
learning.
In order to determine whether single measures of cooperative, competi-
tive, and individualistic interdependence could be derived from the items
or whether the liking and valuing components of the factors needed to be
separated into distinct scales, the factor structures for males and females
were examined and a contrast of Alpha and Rao procedures of factor
analysis was conducted. Examination of the factor structures for males and
females within the 7th-grade and 12th-grade samples provides a clearer
picture of the emergence of item subsets. Subsets of cooperation occurred
only for 7th-grade females, subsets of competition occurred for 7th-grade
males and females, and subsets for individualistic efforts occurred for
7th-grade males and females and for 12th-grade males.
 D. W. JOHNSON AND A. A. NOREM-HEBEISEN 257

Contrast of Alpha and Rao factor analyses were made for the responses
of the 12th-graders in Study 4. Patterns of factor loadings were markedly
similar across all analyses when the number of factors was specified at
three. When the numberoffactors wasleft unspecified, the factor loadings
for competitive and individualistic interdependence were markedly similar
across the two methods, but the item subsets in the cooperative inter-
dependence factor found in the Alpha analysis for the 7th-grade females
were apparent in the Rao analysis for this 12th-grade sample. The oth-
erwise generally high comparability of factor structure across these analyses
indicates that the three attitude scales are factorially robust.
The subdivisions of the major constructs into subsets among 7th-graders
raised the issue of developing six measures of social interdependence rather
than three. The choice to use three attitude scales was made on the basis of
several considerations: (a) item subsets did not occur consistently across
7th-grade and 12th-grade samples, (b) there was high continuity between
responses of the 7th- and 12th-graders when the subsets of items within
each of the three major constructs were combined, (c) the stability of the
three primary factors was evident in the comparison of Alpha and Rao
factor analyses, (d) indices of internal reliability were more robust when the
subset of items found among 7th-graders were combined, and (e) the
purpose of the studies was to develop single measures of cooperative,
competitive, and individualistic attitudes. The three scales (with the sub-
sets of items indicated) appear in the Appendix.
A final check on the factorial structure of the scales was conducted by
submitting the restricted item pool of 22 items to factor analysis. This
analysis was conducted for Studies 4 and 5, as the Ss in these studies had
answeredthe full range of 32 items. As a test of the relative independence
of the three constructs, the items were forced into three factors. The
patterns of factor loading for these analyses reflect a robust and consistent
factor structure with item loadings consistent with the logical contentof the
items. In the 12th-grade analysis, however, three of the individualistic
items show notable strong negative loadings on the cooperation scale
(—.50, —.48, —.45), while in the 7th-grade analysis three of the competitive
items show a positive loading on the individualistic factor (.38, .32, .33)
and one competitive item shows a moderate loading on the cooperative
factor (.37). These loadings are another reflection of the tendency toward
item subsets on the three scales among younger students.
Cronbach alphaindices of internal reliability were computed on the three
factors reflecting attitudes toward cooperative, competitive, and individ-
ualistic social interdependence. From Tables 1 and 2 it may be seen that

Copyright © 2001. All Rights Reseved.

258 JOURNAL OF SOCIAL PSYCHOLOGY

TABLE 1
CRONBACH ALPHA INDICES FOR EVOLVING MEASURES OF SOCIAL INTERDEPENDENCE

Cooperative Competitive Individualistic

Study Items Alpha Items Alpha Items Alpha

Entire school District 3 72 3 82 3 74

12th grade, 1975 4 71 4 73 2 62
7th grade, 1975 4 -65 8 79 7 -80
12th grade, 1976 7 84 8 88 7 84
7th grade, 1976 7 84 8 87 7 84
College, 1977 7 84 8 85 7 88

the internal consistency for each scale increased as the development of

items for each of the constructs advanced with succeeding studies. At the
completion of scale development, indices from .84 to .88 for Studies 4, 5,
and 6.
Table 2 indicates that the cooperative and competitive scales are essen-
tially independent from each other, while a negative correlation exists
between the cooperative and individualistic scales (especially for the older
samples) and the competitive and individualistic scales are somewhat posi-
tively related in the 7th-grade samples.

D. Discussion
Three relatively independent self-report scales with substantial internal
reliability were developed to measure attitudes toward cooperative, com-
petitive, and individualistic interdependence between oneself and others in
an educationalsetting. It is hoped that these scales will be of assistance in

TABLE 2
INTERCORRELATIONS AMONG SCALES OF SOCIAL INTERDEPENDENCE

Sample Competitive Individualistic

Cooperative
Entire school district -02 —.32
12th grade, 1975 68 -.21
7th grade, 1975 wll -.21
12th grade, 1976 .00 —.48
7th grade, 1976 —.03 —.28
College, 1977 —.01 —.60
Competitive
Entire school district 24
12th grade, 1975 .00
7th grade, 1975 22
12th grade, 1976 .04
7th grade, 1976 32
College, 1977 .04
 D. W. JOHNSON AND A. A. NOREM-HEBEISEN 259

future research on social interdependence and educational processes and

outcomes as previously there has been a marked lack of such measures.
The measures developed and their underlying conceptualization are aimed
at American society and do not have anything approachingcultural univer-
sality.
The developmentof the scales provided an ooportunity to examine their
theoretical relationship. The results indicate that the same student may
have positive or negative attitudes toward both cooperation and competi-
tion. Since both cooperative and competitive situations involve interaction
with other people, it may be assumed that a person who is high on both
will be a highly social person wholikes to interact with others in a variety
of ways, while a person whois low on both will generally be a social isolate
who wishes to avoid other people no matter whatthe situation. The results
also indicate that there is a negative correlation between the cooperative
and individualistic scales. While wishing to study alone is obviously nega-
tively correlated with wishing to study cooperatively with others, the size
of the negative correlation between these two scales is not high enough to
support the view that they are opposite ends of a single dimension. It is
also logical that there is some relationship between the competitive and
individualistic scales, as many competitive activities in schools are per-
formed alone. The results indicate that the three types of social interdepen-
dence are not a single dimension, and further research is neededto clarify
the nature of their relationship.
Theresults of the studies reported in this article are congruent with the
position that people’s conceptions of cooperative, competitive, and individ-
ualistic interdependence change as they develop cognitively and socially.
Thescales are more cohesive for older than for younger students, and the
older students become, the greater the negative relationship between the
cooperative and individualistic scales and theless the relationship between
the competitive and individualistic scales. Ahlgren and Johnson (1) found
that there is a reorganization of cooperative and competitive attitudes
during the junior high school years. Thefindings in this study support such
a view. Future research should clarify further the nature and process of
such cognitive change.
APPENDIX
Social Interdependence Scales

a. Cooperative Interdependence.
(1). Liking to cooperate.
T like to help other students learn.

Copyright © 2001. All Rights Reseved.

260 JOURNAL OF SOCIAL PSYCHOLOGY

I like to share my ideas and materials with other students.

I like to cooperate with other students.

(2). Valuing cooperative learning.

I can learn important things from other students.
I try to share my ideas and materials with other students when I think
it will help them.
Students learn lots of important things from each other.
It is a good idea for students to help each other learn.

b. Competitive Interdependence.
(1). Liking to compete.
I like to do better work than other students.
I work to get better grades than other students do.
I like to be the best student in the class.
I don’t like to be second.

(2). Valuing competitive learning.

I like to compete with other students to see who can do the best work.
I am happiest when I am competing with other students.
I like the challenge of seeing whois best.
Competing with other students is a good way to work.

c. Individualistic Independence.
(1). Liking to study alone.
I don’t like working with other students in school.
I like to work with other students. (reverse)
It bothers me when I have to work with other students.

(2). Valuing individualistic learning.

I do better work when I work alone.
I like work better when I ¢o it all myself.
I would rather work on school work alone than with other students.
Working is small groups is better than working alone. (reverse)

REFERENCES
1. AHLGREN, A., & JoHNSON, D. W. Sex differences in cooperative and competitive
attitudes from the second through twelfth grades. Devel. Psychol., in press.
2. DeEuTscH, M. A theory of cooperation and competition. Hum. Relat., 1949, 2, 129-352.
 D. W. JOHNSON AND A. A. NOREM-HEBEISEN 261

3. ——-. Cooperation and trust: Some theoretical notes. In M. Jones (Ed.), Nebraska
Symposium on Motivation. Lincoln, Neb.: Univ. Nebraska Press, 1962. Pp. 275-320.
4. Jounson, D. W. Evaluating affective outcomes of schools. In H. Walberg (Ed.), Evalua-
tion School Performance. Berkeley, Calif.: McCutchan, 1974.
5. Jonnson, D. W. & JoHNSON, R. Learning Together and Alone: Cooperation, Competi-
tion, and Individualization. Englewood Cliffs, N.J.: Prentice-Hall, 1975.
6. ~~, Social interdependence in the classroom: Cooperation, competition, and indi-
vidualism. J. Res. Devel. in Educ., 1978. 12 (Fall).
7. MINNESOTA HIGH SCHOOL TESTING PROGRAM. Minnesota School Affect Assessment:
Manual. Minneapolis: Univ. Minnesota, 1977.

Department of Social, Psychological, and Philosophical

Foundations of Education
University of Minnesota—Twin Cities
330 Burton Hall
178 Pillsbury Drive S.E.
Minneapolis, Minnesota 55455

Copyright © 2001. All Rights Reseved.

Copyright of Journal of Social Psychology is the property of Taylor & Francis Ltd andits content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users mayprint, download, or emailarticles for individual use.