Peer Interaction and Cognitive Development Author(s): Donald R. Rardin and Charles E. Moan Source: Child Development, Vol.

42, No. 6 (Dec., 1971), pp. 1685-1699 Published by: Blackwell Publishing on behalf of the Society for Research in Child Development Stable URL: http://www.jstor.org/stable/1127578 . Accessed: 20/10/2011 17:09
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Blackwell Publishing and Society for Research in Child Development are collaborating with JSTOR to digitize, preserve and extend access to Child Development.

http://www.jstor.org

AND COGNITIVE PEER INTERACTION DEVELOPMENT

DONALDR. RARDIN
University of Florida

CHARLES E. MOAN
Tulane University School of Medicine
RARDIN, DONALD R.,

from 1 classroomeach of kindergarten, first, second, and third grades were ranked on measures of popularity, cognitive and social development. The measure of social development, developed for this study, consisted of combined ratings from 4 measures: Reason for Friendship, Stability of Friendship, Names Not Known, and Congruencyof Friendships. The 2 measures of physical concept development were conservationand classification.Piaget proposed that, during the grades studied, physical and social concepts are closely related to each other and are parallel in their development. He also proposed that various systems of cognition are becoming interconnected during this period and that peer interaction is a causal force helping to bring about this qualitative change in cognition. 2 hypotheses, based on these formulations,were examined. The first, that peer relationsdevelop in a mannerparallel to the developmentof physical concepts, was significantly supported. Both cognitive and social skills progress from kindergarten through the third grade in a similarmanner. A second hypothesis was that a child's cognitive development would be directly affected by the quality of his peer relations,as judged by popularityrankings.Popularitywas found to be closely related to social development, but its relations to measures of physical concept development were relatively minor.
Author Rardin is now serving in the United States Army, Niiremburg, Germany. Requests for reprints should be sent to Charles E. Moan, Behavioral Assessment Laboratory,Department of Psychiatry and Neurology, Tulane Medical School, 1430 Tulane Avenue, New Orleans, Louisiana 70112.
in ChildDevelop[ChildDevelopment, 1971, 42, 1685-1699.@ 1972by the Societyfor Research ment, Inc. All rightsreserved.]

Development.

CHILD DEVELOPMENT, 1971, 42, 1685-1699.

and MOAN,CHARLES E. Peer Interactionand Cognitive

81 children

CHILD DEVELOPMENT
It has been proposed by many developmental theorists that cognitive processes are highly interrelated and interdependent, while learning theorists generally propose processes which are comparatively independent and whose relatedness is an experiential function. The present study primarily concerns the former position and, in particular, cognitive development as proposed by Jean Piaget. In Piaget's system, structure is seen as interposed and mediating between content and function. Briefly, function refers to the manner of cognitive progress; content refers to external behavior; and, structure refers to inferred organizational properties which explain why particular content occurs (Flavell 1963). Cognitive progress is possible through the interaction of assimilation and accommodation. Accommodative acts are always being extended to new and different features of the surround; and, to the extent that a newly accommodated feature can fit somewhere in the existing structure, it will be assimilated. Through assimilation, the structure is changed and further accommodative extensions are possible. Assimilation and accommodation have a cyclical relationship recurring at each level of cognitive development, at first undifferentiated and then progressively separated and coordinated. The present investigation examines structure changes which Piaget infers to take place at about 7 years of age. He proposes that peer interaction is a major factor influencing these cognitive changes and that the social development of the child progresses in a manner parallel to and reflective of cognitive development (Piaget 1952). Accordingly, cognitive development takes place as self and world become differentiated and the accompanying coordination of assimilation and accommodation lead to a state of objectivity and equilibrium (Piaget 1954). With such differentiation, there is a simultaneous process of the establishment of external reality and an inward process of self-awareness. Therefore, ontogenetic development may be viewed as a series of equilibrium (differentiation) achievements, applying not only to sensorimotor acts but also to symbolic manipulations, with the development of cognition at several levels simultaneously proceeding inward and outward from the boundary of self and milieu and the objectification of reality and the reduction of egocentrism proceeding parallel to the acquisition of self-perception (Flavell 1963; Piaget 1954). Piaget has partitioned cognitive development and its structure evolution into three major stages: sensorimotor, concrete operations, and formal operations. The period of concrete operations contains two major substages: (a) preoperational representations (ages 2-7) and (b) concrete operations (ages 7-11). The present investigation concerns the transition from preoperational representations to concrete operations and the assumption that cognition at all levels is the application of actions to the milieu, these actions becoming progressively internalized and less external and observable. 1686

DONALDR. RARDIN AND CHARLES E. MOAN

For example, the concrete operational child has an integrated system from which to operate, with the result that the contents of his world, through the imposition of his structure upon them, are placed in a perspective which takes into account the various facets of the system (Inhelder & Piaget 1958). This child then behaves for the first time in terms of an integrated network of possible actions, with the world beginning to be lawfully organized and thought achieving primacy over perception. Moreover, on the plane of representational thought, social relationships or coordination among individual minds evolves in the same manner as do space relations or object relations. Therefore, it is with the advent of concrete operations that it becomes possible for the child to objectively structure relationships among classes, relations, and numbers, while socially he acquires skill in interindividual relations in a cooperative framework. The acquisition of social cooperation and the structuring of cognitive operations are, then, two aspects of the same developmental processes. The preoperational child, on the other hand, is most prominently characterized by his social egocentrism. Such egocentrism with its lack of introspection, its rigidity, and its lack of logical justification is not overcome simply by experience with objects and events but primarily through social interaction with peers. In the course of such interaction, especially conflicts and arguments, the child is forced to examine his own contentions (Piaget 1928). Therefore, experience, especially with peers, is needed to break down the egocentrism of the preoperational child in order to allow him to consider other perspectives. This "decentration" enables thought to escape the immediate and obvious and encourages the development of conceptual thinking derived from laws and commonalities, as in all true generalization. And, one of the most primary sources of such common laws is interpersonal cooperation and its verifying judgments. The present investigation focuses upon peer exposure and socialization as they relate to cognitive development. Peer exposure is considered the independent variable; and socialization, directly related to the effects of popularity, the dependent variable. In addition, since social development is considered an integral part of cognitive development, their courses of progress should be parallel. Therefore, if Piaget's proposals are correct, progress into concrete operations should vary directly with the amount of peer exposure a child experiences. Further, as Piaget considers cognitive and affective reactions dissociable and parallel but interdependent, the more affectively positive the child's peer interactions are, the more he will seek and the greater will be his progress toward mastery of concrete operations. This study assumes then that the more popular children will receive more peer exposure. Direct support for the hypothesis of a relationship between popularity and cognitive development comes from a study by Goldschmid (1968) in which she found popular peers to be more adept at conservation, while other 1687

CHILD DEVELOPMENT investigations found emotionally disturbed and institutionalized children to be slower in achieving conservation (Goldschmid 1967) and slower at overcoming egocentrism (Neale 1966). Moreover, the predicted trend of social development is also reported in studies not conducted for the purpose of testing Piaget's proposals. Horrocks and Buker (1951), Horrocks and Thompson (1946), and Thompson and Horrocks (1947) demonstrated the trend of greater stability of friendships as children advance in school. Further, while Challman (1932) was unable to establish that friendships among preschool children were based on similarities of the children involved, Furfey (1927) found that similarities such as age, height, weight, and intelligence determine the selection of school "chums." These studies demonstrate some progress toward friendship based on autonomy and differentiation. In addition, Dymond, Hughs, and Raabe (1952) found that second graders place more value upon externals for friendship, while sixth graders more often emphasize personality factors. Such is congruent with the establishment of peer autonomy and with rising above field dependency and surface characterization, as older children should do. A later study (Byrne & Griffitt 1966) found that similarity of attitudes and consensual validation help to determine interpersonal attraction in fourth graders. This result is consistent with Piaget's propositions, and it would be expected that this effect would be observed less often in younger children. Thus, since relevant studies indicated support for the expected course of social development in relation to cognitive development, the following hypotheses are offered for investigation: 1. The development of peer relations and interactions, as reflected by carefully designed and constructed measures of popularity and social decentration will, through the grade range (kindergarten through third grade) tested, parallel that of physical concept development. 2. The individual's progress in cognitive development, as reflected by two measures of physical concept development which assess the ability to classify and conserve, will vary directly with the affective quality of his peer relations. METHOD

Subjects
Eighty-one children (20 kindergarten, 19 first, 21 second, and 21 third graders) in a school in a relatively small, homogeneous Wyoming town were Ss. Of these, there were 80 Caucasians, one Indian; 43 males, and 38 females.

Procedure
Each child was individually assessed on two cognitive measures of 1688

DONALDR. RARDIN AND CHARLES E. MOAN

physical concept development to evaluate the ability to classify and conserve and presented with a task which resulted in a modified sociogram measuring social decentration and popularity.

Cognitive Measures
The first measure of cognitive growth was Form A of the Goldschmid and Bentler Concept Assessment Kit, assessing a child's ability for conservation (the ability to realize that properties such as weight, volume, etc., remain constant despite transformations such as shape or position). Form A consists of six conservation problems which include: two-dimensional space, number, substance, weight, and continuous and discontinuous quantity. The manual of the kit explains in detail the psychometrics of the test. The second measure of cognitive development concerns inclusion or classification skills and follows a procedure developed by Kofsky (1966), who used tests which demonstrated six different, ordered levels of difficulty of classificatory logic. Three of these steps, occurring during the age levels of the present study and employed as measures of cognitive progress, were: (a) "a grasp of the elementary relations among objects and classes in a hierarchy including the knowledge that an object can belong to more than one class . . . which was demonstrated by . . . a majority of the six-yearolds"; (b) "conservation of hierarchy . . . and reclassification, which were performed successfully by most of the seven-year-olds"; and (c) "knowledge of inclusion which was apparent in most of the nine-year-olds." The actual tasks involve the ability to correctly manipulate a set of geometric blocks whose plane surfaces were either square, circular, or triangular and whose colors were red, blue, green, or yellow. The administration was the same as presented by Kofsky except that all tasks were presented in the same order to each S.

Modified Sociogram
This measure, a sociometric status procedure similar to that employed by McCandless (1957), with a retest after 6 weeks, involved presenting each child a picture of his class and asking him to show which three children he liked most (best) at that time. The procedure was modified and supplemented by questions attempting to ascertain reaction to present field, the basis of friendship in terms of reciprocity, and individual autonomy. Immature socialization, relative to other children, would be represented by: instability of friendships (based on retest); lack of interpersonal awareness, or egocentricity (based on how well the child is aware of others' feelings); and lack of individual autonomy (determined by basis of friendship and whether it is reciprocal). After each child was individually presented a picture of his class, the following inquiries were made for scoring purposes: 1689

CHILD DEVELOPMENT
This is a picture of your classmates, do you recognize them? Can you show me some of your friends? Can you show me the three friends you like the best? Why? Which one of those do you like the very best? Why? How long have you known him (or her)? Where did you meet him (or her)? Can you name all the children in the picture? Which boy do you think most of the kids like best? Which girl do you think most of the kids like best? The question "Can you show me the three friends you like best?" was repeated in a retest after 6 weeks. In both cases, S was to identify three individuals who were actual best friends, not three whom he would like to have as best friends.

Scoring
The conservation and classification measures were scored by ranking Ss once for the number of errors on the conservation task and once for the number of errors on the classification task. For ties, a mean rank was assigned for each group of tied individuals. The popularity measure was obtained by counting how often a child was chosen by Ss as one of the three children in the class they liked best. Scores were assigned to each S according to how often he was chosen and were ranked as with the other measures. From the questions concerning socialization, each S was ranked five times, using the following criteria: (1) reason for choosing best friend (Reason); (2) stability of friendships (Stability); (3) numbers of names of the children in the class that were not known (Names Not Known); (4) matching of friends chosen with each other (Congruency); and (5) ability to choose children which other children like best (Popularity Perception). The scoring methods for these five measures are as follows: 1. Reason scores were achieved by grouping reasons given by children into three categories: impersonal, personal, and interpersonal. The impersonal category was usually a fact of circumstance such as a cousin or living on a farm. The personal category includes personal actions such as playing with another child. The third category includes reasons which show awareness of the other child, for example, "He's nice," or "He's fun to play with." The Ss were given scores of 1, 2, or 3 depending upon into which group they were placed. 2. For Stability scores, Ss were grouped according to the number of friends on retest which were the same as on initial testing, 0, 1, 2, or 3. 3. Scoring for Names Not Known was achieved by grouping Ss according to number of children in their individual classes whose names were not known and assigning a score equal to that number, corrections being made for small differences in classroom size. 1690

DONALD R. RARDIN AND CHARLESE. MOAN 4. Congruency scoring was obtained by counting the number of matches between an individual's six choices of best friends (three on initial testing and three on retest) and their reciprocal choice of him as a best friend. The individuals were then scored according to the number of matches (0--6). 5. Popularity Perception scores were achieved from the answers to the questions concerning which boy and which girl S thought the other children in the class liked best. A S's choices were compared with the actual popularity rankings. Rankings then proceeded according to differences between the most popular and the one perceived as most popular for each S. The highest ranking was accorded to Ss whose choices for being most popular were in fact the most popular. It was planned to rank Ss according to how they stated they met the child whom they liked the best, but almost every child replied that they simply met at school. The same applies to length of friendship. Therefore, these criteria were not included in data analyses. In each of the five scored categories, Ss were ranked according to their scores, a mean rank being assigned to tied individuals. RESULTS The Kruskal-Wallis test, a nonparametric "analysis of variance by ranks" (Hays 1963), compared the population distribution of the various rankings (popularity measure, conservation task, classification task, and the five measures of socialization) across grades (kindergarten, first, second, and third), as is shown in table 1, where H is referred to a X2distribution with 3 df (Hays 1963). All measures except Stability are significant, suggesting that the distributions of the various measures are not the same for each group. Stability approached significance with p < .25. Further, the comparison of grade means (table 2) demonstrates progressive gains from kindergarten through TABLE 1
KRUSKAL-WALLIS COMPARISONS OF POPULATION DISTRIBUTION FOR ALL RANKINGS ACROSSGRADE LEVEL

Measures Classification ......................... Conservation ......................... Stability ............................. Names Not Known ................... Reason .............................. Congruency .......................... Popularity Perception .................. **P < .05.
"p < .01.

Rankings H= H= H= H= H= H.=H H = 34.52*** 29.18*** 4.95 22.19*** 16.77*** 9.49** 24.80***

1691

CHILD DEVELOPMENT
TABLE 2
MEAN RANKS FOR EACH MEASURE AT EACH GRADE LEVEL

Classification .................. Conservation .................. Stability ...................... Names Not Known ............ Reason ....................... Congruency ................... Popularity Perception ..........

63.0 63.8 51.1 62.2 56.1 47.9 54.1

46.6 41.7 37.9 36.3 43.0 48.3 23.7

34.4 33.4 37.6 35.2 39.7 40.3 32.1

21.6 26.2 37.5 30.9 26.0 28.6 53.0

the third grade for all means except Popularity Perception and Congruency, although the Congruency measure progresses essentially as expected. All other rankings become progressively higher with successive grades (the rank of 1 being the highest possible). In order to obtain a measure of socialization, the rankings for Stability, Names Not Known, Reason, and Congruency were combined into a single measure. The four ranks for each individual were summed, and the resulting combined scores were themselves ranked (table 3). Popularity PercepTABLE 3
MEAN RANKS OF THE SOCIALIZATION MEASURE FOR EACH GRADE AND COMPARISON OF ADJACENT GRADE-LEVEL MEANS FOR EACH COGNITIVE MEASURE

K Socialization ............... 64.20

1 41.47

2 36.26

3 23.21

GRADES COMPARED

K-1 Conservation .................. Classification .................. Socialization ................... 3.13*** 2.21** 3.06***

1-2 1.17 1.65** .67

2-3 1.04 1.78** 1.80***

NOTE.-For Z scores, ** p < .05 and *** p < .01.

tion was not included, since it was obviously measuring something different from the other four measures. As can be seen from the mean scores for socialization, conservation, and classification, the measures progress in a similar manner from grade to grade. Figure 1 illustrates this similarity. For each measure, there is a progressive improvement from kindergarten through the third grade. Dunn's (1964) technique of "Multiple Comparisons Using Rank Sums" demonstrates that the majority of differences between successive grades, for each measure, reach significance (table 3). When classes which are two grades apart, rather than one, are .compared, the differences are all significant. 1692

DONALDR. RARDIN AND CHARLES E. MOAN

70
----CONSERVATION
CLASSIFICATION

60Z 50

........ SOCIALIZATION

4=

\4

<

4030 20

GRADE
FIG. 1.-Mean ranks of the three cognitive measures for the four grades

Kendall's T with ties, employed as a rank-order correlation method, was used to examine the relations among conservation, classification, and socialization measures (Hays 1963). Table 4 shows the significance of the relationships among the three cognitive measures. All coefficients are significant at the .01 level. This significance, however, may be artificial, because comparisons of the three measures with grade in school also have very significant T coefficients (table 4). While this again shows the similarity in function of the three measures, it also suggests that the correlation coefficients of their interrelations may only reflect their mutual relations with grade in school. If the rank-order correlations among measures are considered within grade level rather than across grade level, the levels of correlation are much smaller, as seen in table 5. Here, the coefficients among measures are smaller, and only in the first grade do they consistently reach significance. For the first-grade children, socialization correlates with a coefficient of .25 with conservation and with a coefficient of .31 with classification (significance at .05 and .01, respectively). 1693

DEVELOPMENT CHILD
TABLE 4
KENDALL'S T COEFFICIENTS FOR THE THREE COGNITIVE MEASURES AND FOR THE COMPARISON OF COGNITIVE MEASURES WITH GRADE IN SCHOOL

Coefficient

........................ Socialization-conservation Socialization-classification ........................ ........................ Conservation-classification Socializationand grade .......................... Classification and grade .......................... Conservationand grade ..........................
*** p < .01.

.26*** .39*** .47*** .48*** .47*** .50***

Socialization appears to function similarly to conservation and classification throughout the preceding statistical procedures. However, their actual relationships to each other are relatively minimal when measures are compared within each grade. Kendall's - was also used to examine the hypothesis that cognitive development, as measured, would vary directly with popularity. Table 5 lists the rank-order coefficients for popularity compared with each of the cognitive measures in each of the four grades. The levels of significance were determined by z scores corrected for continuity (Hays 1963). As can TABLE 5
KENDALL'S

T COEFFICIENTS

FOR RELATIONS AMONG COGNITIVE MEASURES WITHIN GRADES AND WITHIN-GRADE COEFFICIENTS FOR POPULARITY RANKINGS AND THREE COGNITIVE MEASURES COGNITIVE MEASURES

GRADE LEVEL

Socialization and Conservation .046 .25** - .017 .37***

Classification and Conservation .31** .36*** .21 .10

MEASURE WITH POPULARITY

Socialization and Classification .035 .31*** .080 - .081

K 1 2 3

.................... .................... .................... ....................

Socialization K 1 2 3 .................... .................... .................... ....................

*p<.1o. ** < .05. *** p p < .01.

Conservation .13 .29* .07 .16

Classification .27* .13 .05 .15

.11 .60*** .47*** .25*

1694

DONALD R. RARDIN AND CHARLESE. MOAN be seen, all the coefficients are in the predicted direction; but the relation of popularity to conservation or classification appears to be relatively minor, while the relation between socialization and popularity is consistent and reaches significance. During the first grade, the correlation coefficient reaches a high of .60. In summary, the hypothesis that peer relations develop in a manner parallel to the development of physical concepts was significantly supported. Both cognitive and social skills were found to progress from kindergarten through third grade in a similar manner. The second hypothesis was that a child's cognitive development would be directly affected by the quality of his peer relations, as judged by popularity rankings. Popularity was found to be closely related to social development, but its relations to physical concept development were relatively minor. DISCUSSION As hypothesized, it was found in the present investigation that the measure of socialization and measures of physical concept development (conservation and classification) increased progressively from kindergarten through the third grade and showed very significant differences between and extremely similar mean ranks within grades. The socialization measure was derived from a combination of four others (reason for friendship, stability of friendships, names of children in class not known, and reciprocity or congruency of friendships). Popularity perception had been planned as a fifth inclusion, but it was found that first-grade Ss received a higher mean ranking on popularity perception than kindergarten Ss and that second and third graders received progressively lower rankings than did first graders. These variable, unexpected results may be a function of children becoming more demanding of reciprocity of friendship as they become more adept in interpersonal perception. Thus, while there is much agreement as to which children are liked best in the first grade, this agreement declines in second and third grades, possibly because the first-grade children who were so popular cannot later be reciprocally close friends with all the children who like them. Therefore, in the later grades, "best friend" choices are spread out over more children. The measure of popularity in the present study did not reflect this interaction adequately for inclusion in the measure of socialization. The second hypothesis, that cognitive progress varies with the quality of S's peer relations, is not clearly supported by the data. Rank-order correlations are quite significant when popularity and socialization are considered; but, when the rank-order correlations between popularity and conservation or classification are considered, only two of six resulting coefficients show a significant, but minimal, result. Indirect effects of popularity by way of its effects upon socialization might be derived from 1695

CHILD DEVELOPMENT
correlations among socialization and conservation and classification. The correlations among the three measures themselves are highly significant if taken across grades, but significant results are primarily limited to the first grade if the correlations are taken within grades. The three measures all correlate highly with grade in school; and the relationship among the measures, across grades, may result from each of their relationships with age. Therefore, social awareness appears significantly relat'ed to this measure of popularity, but neither popularity nor socialization has an important relation with the development of physical concepts. This latter statement, however, must be modified in view of the z correlations of .25 between socialization and conservation and of .31 between socialization and classification (significant at the .01 and .05 levels, respectively) -both at the first-grade level. There are other considerations possibly affecting these correlations. First, administrative differences might be important, since the three measures of this study obtained their information in different manners: (a) the conservation measure contained forced-choice questions; (b) the classification measure contained both forced-choice and open-ended questions; and, (c) the socialization questions simply solicited information rather than presenting problems and asking questions. Further, only in the first grade is there a wide distribution of scores within each measure-particularly the measure for conservation. In kindergarten, nearly all Ss miss every item of the conservation measure; and, already in the second grade, Ss are successfully answering most items. Thus, the comparisons of socialization and classification measures, which appear to span a broader grade and age range, with conservation are possibly distorted; because, at any but the first grade, they are being compared with the extreme ends of the conservation measure, where there is a very narrow range of difference. It would be expected, therefore, that the most likely grade for significant correlations would be the first, as is the case. The importance of these correlations is that Piaget proposed that the various cognitive skills gained at this level reflect an underlying whole which, in part, determines these cognitive changes. The learning principles of behaviorism would predict more independent development. An argument might be put forth that only in the first grade are the correlations among measures significant and that even then the coefficients are small. An answer in support of a Piagetian view would be that, first, the significant results are found only at the first grade, which may be partially an artifact of the measures employed, and that, second, the interrelations among socialization, conservation, and classification should decline after about the age of 8. These are circumscribed skills which can be learned in their totality at this age and not open-ended skills which would be continually extended for years. Therefore, the independent variables and the relations among the skills themselves will function as such only while the skills are coming into being. 1696

DONALDR. RARDIN AND CHARLES E. MOAN

The hypothesis that cognitive development will vary directly with the affective quality of peer relations receives varying amounts of support, depending upon how the question is approached. If the present measure of socialization is accepted as a reflective measure of cognition (which the data concerning the first hypothesis appears to support), then, at least on this particular measure, the second hypothesis receives significant support with the following correlations between popularity and socialization: kindergarten, .11; first grade, .60 (p < .01); second grade, .47 (p < .01); and, third grade, .25 (p < .10). Direct evidence for the interrelationship of popularity with either conservation or classification was limited. Grade level appears to show much greater influence as a common denominator with Kendall's r correlations between grade and cognitive measures, being .48, .47, and .50 for grade and (a) socialization, (b) classification, and (c) conservation, respectively (all significant at .01). The closeness of these figures also lends further support to the first hypothesis. There are three major areas of implication in this study: childdevelopment theory, education, and psychotherapy. In the area of child development, there does indeed appear to be a qualitative change at the second grade, about 7 years of age, in cognitive processes. This change demonstrates many of the qualities which Piaget described as grouping of structure. Grade was the strongest common factor with the three measures in the present study. Popularity appeared significantly related to the three measures in the first grade, but, in general, its effects on conservation and classification were slight. Two possible conclusions would be that either (1) the popularity measure was not indicative of the amount of peer interaction a child experiences or that (2) the amount of peer interaction, which forces a child to examine his own contentions, affects physical concept development in only a minor capacity. It is presently felt that the popularity measure was consistent with Piaget's proposals and that the latter conclusion coincides with both the present data and Piagetian theories. The former conclusion would be open to further experimental investigation. In the area of education, it appears that the best way to learn awareness of others is by having active exposure with them. If, in fact, this is a learned skill, it is an area which formal education is often accused of hampering. Highly structured, noninteractive classes during the early ages would appear to discourage this type of growth, with such also holding true at later ages and different levels of cognitive development. On the other hand, currently popular exercises such as "show and tell" would appear to encourage individual autonomy and awareness, and schools of the prototype of Summerhill would appear to develop socialization to its furthest degree. However, it also appears that the effects of peer interaction upon other areas of cognition is limited; for learning these concepts, it seems that adult-mediated direct exposure to them would be necessary. 1697

CHILD DEVELOPMENT Finally, in the area of psychotherapy, it would appear that group therapy before age 6 would have few, if any, therapeutic advantages. During the ages when concrete operations are being gained, a time when many children are seen in clinics because of difficulties experienced in adapting to the pupil role, careful consideration might be given to the use of peer groups in which peer interactions were encouraged. The socialization process, which peer interaction appears to aid (from data in the present experiment), would seem to be a benefit to such children; and the social skills learned at this age would obviously be very useful with regard to the child fitting into a classroom situation. REFERENCES Byrne, D., & Griffitt, W. A developmental investigation of the law of attraction. Journal of Personality and Social Psychology, 1966, 4, 699-702. Challman, R. C. Factors influencing friendships among preschool children. Child Development, 1932, 3, 145-158. Dunn, O. J. Multiple comparisonsusing rank sums. Technometrics, 1964, 6 (3), 241. Dymond, R. F.; Hughs, A. S.; & Raabe, V. L. Measurable changes in empathy with age. Journal of Consulting Psychology, 1952, 16, 202-206. Flavell, J. H. The developmental psychology of Jean Piaget. Princeton, N.J.: Van Nostrand, 1963. Furfey, P. H. Some factors influencing the selection of boys' chums. Journal of Applied Psychology, 1927, 11, 47-51. Goldschmid, M. L. Different types of conservation and nonconservationand their relation to age, sex, IQ, MA, and vocabulary. Child Development, 1967, 38, 1229-1246. Goldschmid, M. L. The relation of conservation to emotional and environmental aspects of development. Child Development, 1968, 39, 579-589. Hays, W. L. Statistics for psychologists. New York: Holt, Rinehart & Winston, 1963. Horrocks, J. E., & Buker, M. E. A study of the friendship fluctuations of preadolescents. Journal of Genetic Psychology, 1951, 78, 131-144. Horrocks, J. E., & Thompson, G. G. A study of friendship fluctuations in rural boys and girls. Journal of Genetic Psychology, 1946, 69, 187-198. Inhelder, B., & Piaget, J. The growth of logical thinking from childhood to adolescence. New York: Basic, 1958. Kofsky, E. A scalogram study of classificatorydevelopment. Child Development, 1966, 37, 191-204. McCandless, B. R. Peer popularity and dependence on adults in preschool-age children. Child Development, 1957, 32, 511-518. Neale, J. M. Egocentrism in institutionalized and noninstitutionalized children. Child Development, 1966, 37, 97-101. Piaget, J. Judgment and reasoning in the child. New York: Harcourt, Brace, 1928. 1698

AND CHARLES E. MOAN DONALDR. RARDIN

Piaget, J. The origins of intelligence in children. New York: International Universities Press, 1952.

Piaget, J. The constructionof reality in the child. New York: Basic, 1954.
Thompson, G. G., & Horrocks, J. E. A study of friendship fluctuations among urban boys and girls. Journal of Genetic Psychology, 1947, 70, 53-63.

1699