Effects of Age, Sex, and Physical Attractiveness on Child–Peer Relations, Academic Performance, and Elementary School Adjustment
Richard M. Lerner and Jacqueline V Lerner
A notion from organismic developmental theory is that on the basis of a child’s possessing behavioral (e.g., temperamental) and/or physical (e.g., constitutional) characteristics of individuality, differential reactions may be evoked in significant others, and these can feed back to affect further development (Lerner, 1976). Through the establishment of such “circular functions” in ontogeny (Schneirla, 1957), a child’s own characteristics may provide a significant source of stimulation for his/her own development.
Interpersonal popularity data derived from sociometric responses of nursery school children (Dion & Berscheid, Note 1) support the idea of such a circular relation between characteristics of physical attractiveness and educational, social, and personality developments. By the end of the preschool term unattractive children were liked less by their peers than were more attractive children. Moreover, consistent with adult physical attractiveness stereotypes (see Berscheid & Walster, 1974), the children believed that aggressive, antisocial behavior (e.g., hitting and yelling at the teacher) was more characteristic of the unattractive children, while attractive children were seen as more independent, appeared not afraid of anything, and were seen as self-sufficient. Thus, attractive children appear perceived as more, and unattractive children as less, adjusted to people and activities involved in the school situation.
In addition, evidence indicates that, as do children, adults believe that attractive children engage in more socially desirable behaviors than do unattractive children. Dion (1972) found that college women evaluated an unattractive child’s transgression as being likely to be repeated and as indicative of a chronic antisocial character, whereas these were not the interpretations offered when the identical transgression was committed by an attractive child. Thus, differential peer and adult appraisals may provide an experiential context channeling the child toward the emission of adjustment-related behaviors congruent with expectation. Simply, a self-fulfilling prophecy/ circular reaction may exist, and this may affect the child’s educational interactions. In fact, Clifford and Walster (1973) found that experienced fifth-grade teachers moderated their professional evaluation of a student in relation to his/her physical attractiveness, and again, with a favorable bias towards the attractive.
In sum, evidence suggests that, compared to the physically attractive child, the unattractive child experiences negative and rejecting peer relations, the perception of maladjustment by both teachers and peers, and the belief by teachers of less educational capability. Development in such an interactional climate might foster the emergence of the very behaviors and characteristics expected by the unattractive child’s socializing others. In short, on the basis of a self-fulfilling prophecy/circular reaction the unattractive child should emit maladjusted behaviors, show less educational achievements and abilities, and enjoy less favorable and more unfavorable peer interactions than does the attractive child. These inferences were tested in this study.
Subjects were 56 fourth-grade children (61% male, mean age = 9.6 years, SD = .4) and 48 sixth-grade children (50% male, mean age =11.8 years, SD = .5). The sample (total n = 104, 94% white) was derived from the only elementary school in a homogeneous, semi-rural, working-class community of approximately 15,000 people in southeastern Michigan. Although most parents in this preponderantly (over 95%) white area are high school graduates, few have any college education, and the predominant vocations (over 85%) are factory worker, clerk/salesperson, and laborer. Subjects were tested in December 1975, and all children present on the days of testing were included as subjects.
Actual and Perceived Academic Performance
Several measures were compiled on each child. Academic achievement scores for the child’s current year in school and for each of the two preceding years were derived from a composite of each student’s marks; these marks were converted into a grade point average for each grade by making A+ = 12, A = 11, A— = 10, and so on through failure = 0 and then averaging these scores. Perceived academic ability was indexed by asking each student’s teacher to rate such ability on a seven-point Likert-type scale, with response alternatives ranging from 1 = ability is extremely above average through 7 = ability is extremely below average.
Actual and Perceived School Adjustment
Actual adjustment was indexed through use of two sections of the Social Adjustment portion of the California Test of Personality; these two sections— Social Skills and School Relations—have alternate-form reliabilities of .84 and .88, respectively. Perceived academic adjustment was indexed by asking each students teacher to rate such adjustment on a seven-point Likert-type scale, with response alternatives ranging from 1 = adjustment is extremely above average through 7 = adjustment is extremely below average.
Positive and Negative Peer Relations
A positive and negative peer relations score was derived for each subject. From a list of positive and negative bipolar phrases (Lerner & Korn, 1972), describing social and personal attributes, nine positive attributes (other boys and girls like him (her), most want as a friend, happy, doesn’t fight, have many friends, be picked leader, picks the games to play, kind, and neat) and nine negative attributes (least want as a friend, gets teased, is left out of games, mean, fights, sloppy, sad, have few friends, and not be picked leader) were selected. Subjects were read the randomly listed items and were asked to name a boy (girl) in their class that best fit each item. A subjects positive or negative peer relations score was thus derived by summing the favorable or unfavorable mentions, respectively, he or she received from his or her peers.
Physical attractiveness ratings were obtained for each subject in two ways. First, the experimenter gave each subject a global physical attractiveness rating, based on her first impression of the subject; her response alternatives ranged from 1 = very attractive to 5 = very unattractive. Second, a photograph was taken of each subject, immediately following the subjects individual assessment. A midchest-through-head color slide was made through the use of a Kodak 28 Instamatic camera. All subjects’ slides were presented to a large group of introductory psychology students (N= 97) through use of a Kodak Carousel Projector. The college students rated each side for physical attractiveness through use of the above five-point scale. College students were used for these ratings because of the finding that physical attractiveness, as judged by adults, is related to a child’s peer popularity (Berscheid & Wal- ster, 1974). Accordingly, the mean of the college students’ ratings for each subject was used as the second index of the child’s physical attractiveness.
Before any individual interaction with the subjects, the two scales of the California Test of Personality were administered to subjects, tested as groups within their classrooms. At this time the teacher rated the adjustment (or ability) of each of her students. On the following school day, she rated each student for the remaining dimension; thus, teacher ratings of adjustment and ability were indexed in counterbalanced order. A teacher did not see her first set of ratings when making the second.
After this first day subjects were individually interviewed to obtain the peer relations scores. The randomly ordered positive and negative items were read to the subject. In counterbalanced order, each subject was asked to name a boy (or a girl) in the class who best fit or most displayed each item. All questions about one sex were completed before questions about the other sex were asked. At the beginning of the interview, before talking to the child, the experimenter told the subject that she wanted a photograph to remember them by. She promised (and delivered) a copy of the photo to each child. This individual session lasted about 5 minutes per child.
Slides (6.4 cm x 6.4 cm) of each subjects photograph were constructed and presented to the 97 introductory psychology students used as adult raters of physical attractiveness. The raters, tested as a group within their classroom, used the five-point scale of physical attractiveness rating, described earlier, to indicate their judgments. The 27 males and 70 females (mean age = 20.5 years; SD = 2.15 years, 91% white) rated all of the 104 slides in about 15 minutes.
The means and standard deviations of all variables, and the product-moment correlations among them,1 are presented in Table 2.1. Grade point average was determined for all subjects for one and two years preceding their current year in school and for their current year. As seen in the table this average remained stable across these years, and the index of perceived academic ability, obtained through teacher ratings, was highly related to these actual academic achievements. As also seen in Table 2.1, high consistency was found between the teachers’ ratings of student academic adjustment and actual achievement. However, although a high relation obtained between the teachers’ ratings of perceived academic ability and perceived academic adjustment, there was little relation between perceived adjustment and the Social Skills and the School Relations subtests of the California Test of Personality.
Data in Table 2.1 also indicate support for the hypothesis that the positive and the negative peer relations scores would be related to physical attractiveness. A positive relation obtained between positive peer relations and both the experimenter’s and group mean rating of physical attractiveness, while negative relations were found between negative peer relations and both these measures of physical attractiveness. The hypothesis that high physical attractiveness predicts positive interactions not only with and favorable appraisals by peers but with and by adults too, while the reverse pattern of
relations should obtain with low attractiveness, is also supported. Teachers’ academic and adjustment ratings were significantly related to both measures of physical attractiveness. These teacher appraisals were related also to both the positive and negative peer relations scores.
It has been predicted, on the basis of a presumed circular function/self- fulfilling prophecy feedback process, that the teacher perceptions and peer appraisals should relate to actual behavior emitted by children. As also seen in Table 2.1, some support for this prediction exists, at least insofar as patterns of covariation with the measures of academic ability are concerned. Both present grade point average and grade point average two years ago were related to physical attractiveness as indexed by both measures; grade point average one year ago bore a similarly directed, but marginal relation to these two physical attractiveness measures (p < .10 in both cases). Moreover, there is tentative evidence for a relation between actual adjustment, as indexed by the two California Test of Personality subscales, and physical attractiveness. Although Social Skills subtest scores were unrelated to either the group mean attractiveness rating or to the experimenter’s attractiveness rating, marginally significant relations between School Relations scores and these two measures obtained (p < .09 in both cases). Finally, neither grade level nor sex of subject were related to any of the measures.
Analyses of Variance
To test whether subtle interactions exist among physical attractiveness, grade, and sex at particular levels (e.g., extremes) of attractiveness, the more reliable group mean ratings of physical attractiveness were used as the basis of dividing subjects on the basis of physical attractiveness scores. Within each grade, three physical attractiveness groups were formed. Students whose mean rating was in the highest quartile (range = 1 to 2.39) were placed in a high physical attractiveness group, students whose mean rating was in the middle two quartiles (2.4 to 2.89) were placed in an average group, and students whose mean rating was in the lowest quartile (2.9 to 5) were placed in a low attractiveness group. After this placement nine 2x2x3 fixed effects analyses of variance were performed, with the dimensions consisting of grade (fourth or sixth), sex, and physical attractiveness group (high, average, low), respectively. All factors in these analyses were between-subject dimensions. Because of the presence of unequal cell ns, unweighted means approximations (Winer, 1971) were done. Successively analyzed dependent measures were the first nine variables listed in Table 2.1.
Consistent with the correlational analyses, no significant main effects for grade or sex were found, and no two- or three-way interactions obtained in any analysis. Main effects were found only for physical attractiveness level and only with four of the analyses. All findings were consistent with expectations. Specifically, when teachers’ academic ability ratings was the dependent variable, a main effect for physical attractiveness was found, F(2,92) = 3.15,
p < .05, indicating that teachers rated more attractive children as having higher academic abilities (M = 2.4) than those children of average (3.3) or low (3.9) attractiveness. The main effect for physical attractiveness found when teachers’ academic adjustment ratings was the dependent measure, F (2,92) = 7.18, p < .002, indicated that the teachers perceived the more attractive children as being better adjusted to the educational environment (M = 2.4) than children of average (3.2) or low (3.8) attractiveness. When positive peer relations scores was the dependent variable, the effect for physical attractiveness, F (2,92) = 2.54, p < .08, suggested that the high (16) and moderately (14.1) attractive children enjoyed more favorable interactions with peers than did their less attractive peers (5). The last main effect for attractiveness, found when negative peer relations was the dependent measure, F (2,92) = 4.35, p < .03, indicated that less attractive children experienced more negative peer appraisals (22.96) than did children of average (8.01) or high (5.8) attractiveness.
The predicted covariation among children s physical attractiveness, peer and teacher appraisals, and actual behavior functioning were supported, and as such suggest that the psychosocial developmental milieu of the physically attractive child is more favorable than that of the physically less attractive. Although the percentages of variance accounted for by physical attractiveness are not large, averaging about 10%, they tend to be consistent and significant across analyses. Thus, these findings, considered in the context of previous research (Berscheid & Walster, 1974), indicate that a child’s physical attractiveness represents an instance of an organismic characteristic which may provide a significant source of the child’s own development. In other words, although not derived from repeated measures, the present data are consistent with expectations derived from a circular function idea.
Here then lie the theoretical implications of the present findings. Children, differing in physical attractiveness, evoked differing reactions in their peers and adult supervisors. On the basis of such differential reactivity to child individuality, it is argued (Lerner, 1976; Schneirla, 1957) that a child will experience differential feedback and, as a consequence of the progression of such a circular function, further lawful, individual development will proceed. The consistency between the peer and teacher perception ratings and physical attractiveness, on the one hand, and the measures of actual behav- ior/personality and physical attractiveness, on the other, is congruent with such a circular function. This function would involve a linking of appraisals of ability based on physical attractiveness stereotypes with differential treatment based on such appraisals, and would eventually lead to behavior channeling and hence stereotype consistent performance. Thus, the general relations of physical attractiveness to grade point average are consistent with the presence of such an academic circular function/self-fulfilling prophecy.
However, it must be noted that although the present study’s findings are consistent with the organismic circular function feedback notion, they in no way provide unequivocal support of this construct. The present study involved only assessing covariation among numerous indices, largely measured at one point in time and thus cannot provide a direct test of the circular function notion, a concept which necessarily implies a longitudinal assessment. Thus, any adequate test of the interpretative utility of this notion must involve an appropriate repeated measures (i.e., sequential) design (e.g., see Baltes, 1968). This design should include prediction of consequent outcome on the basis of antecedent organismic variable status and the inclusion of experimental manipulations (cf. Schaie & Baltes, 1975) across the sampled ontogenetic span.
Finally, although the findings of the present study indicate that the implications of physical attractiveness seem to be stable across sex, future research should be conducted to determine if this stability remains constant with age. Specifically, although there were no sex differences as regards the role of physical attractiveness for the late-elementary school children assessed in this study, it is possible that perhaps at the time of puberty', with changes in body' shape and the emergence of secondary' sex characteristics, there would be sex differences.
This study' is based in part on a thesis submitted by Jacqueline V. Lerner to the Department of Psychology, Eastern Michigan University, in partial fulfillment of the requirements for the degree of master of science. The second author is now at The Pennsylvania State University. John R. Knapp and Barbara Brackney are thanked for critical readings of an earlier version of the manuscript.
1 Dion, К. K., & Berscheid, E. Physical attractiveness and social perception of peers in preschool children. Mimeographed research report available from E. Berscheid, University of Minnesota, Elliott Hall, Minneapolis, MN.
1. All correlations are based on df= 102.
Baltes, P. B. (1968). Longitudinal and cross-sectional sequences in the study of age and generation effects. Human Development, 11, 145-171.
Berscheid, E., & Walster, E. (1974). Physical attractiveness. In L. Berkowitz (Ed.), Advances in experimental social psychology. New York: Academic Press.
Clifford, M. M., & Walster, E. (1973), The effect of physical attractiveness on teacher expectation. Sociology of Education, 46, 248-258.
Dion, К. K. (1972). Physical attractiveness and evaluations of children’s transgressions. Journal of Personality and Social Psychology, 24, 207-213.
Lerner, R. M. (1976). Concepts and theories of human development. Reading, MA: Addison- Wesley.
Lerner, R. M., & Korn, S. ). (1972). The development of body build stereotypes in males. Child Development, 43, 908-920.
Schaie, K. W., & Baltes, P. B. (1975). On sequential strategies in developmental research and the Schaie-Baltes controversy: Description or explanation? Human Development, 18, 384-390.
Schneirla, T. C. (1957). The concept of development in comparative psychology. In D. B. Harris (Ed.), The concept of development. Minneapolis, MN: University of Minnesota Press.
Winer, B. J. (1971). Statistical principles in experimental design. New York: McGraw-Hill.