The Development of Self-Conceptions from

Developmental Psychology 1977, Vol. 13, No. 4, 314-319

The Development of Self-Conceptions from Childhood to Adolescence R A Y M O N D MONTEMAYOR Department of Family and Consumer Studies The University of Utah

M A R V I N EISEN Michigan State University Self-concept development from childhood to adolescence was studied from a cognitive-structural perspective. The responses of subjects to the question "Who am I?" were analyzed by means of a 30-category scoring system. Between childhood and adolescence, there was a significant increase in self-conceptions categorized as follows: occupational role; existential, individuating; ideological and belief references; the sense of self-determination; the sense of unity; interpersonal style; and psychic style. A decrease occurred for self-conceptions based on territoriality, citizenship; possessions, resources; and physical self, body image. Curvilinear age changes were found for the use of the categories sex; name; kinship role; membership in an abstract category; and judgments, tastes, likes. The results for self-concept development were in general agreement with Werner's notion that cognitive development proceeds from a concrete to an abstract mode of representation.

Traditional approaches to the study of self-concept development between childhood and adolescence have been primarily based on either role theory (Elder, 1968) or a psychodynamic perspective (Erikson, 1968; Freud, 1969). The focus of these theoretical orientations is on possible ontogenetic changes in traits, dispositions, or motivational states (Carlson, 1965; Engel, 1959; Monge, 1973) and on identifying age changes in the factor organization of traits applicable to the self (Kokenes, 1974). One aspect of self-concept development previously neglected concerns possible structural changes in the types of constructs used to describe oneself. Both Piaget (In-

helder & Piaget, 1958) and Werner (1961) have argued that an individual's cognitions about the physical world undergo important qualitative changes between childhood and adolescence. In particular, Werner's "orthogenetic principle" states that "whenever development occurs, it proceeds from a state of relative globality and lack of differentiation to a state of increasing differentiation, articulation, and hierarchic integration" (Werner, 1957, p. 126). Werner's principle suggests that as an individual matures, his cognitions about the physical world undergo a shift from a concrete to an abstract mode of representation. The orthogenetic principle, although not specifically concerned with ontogenetic This research was supported in part by City Univer- changes in social cognition, has been fruitsity of New York Faculty Research Award 10702 to the fully extended by Crockett (1965) to the area first author. of the development of person perception. Marvin Eisen is now at the Institute for Social and According to Crockett, an individual's dePolicy Studies at Yale University. veloping ability to think abstractly allows Requests for reprints should be sent to Raymond Montemayor, Department of Family and Consumer him to differentiate between another perStudies, The University of Utah, Salt Lake City, Utah son's appearance or behavior and his under84112. lying dispositional qualities. Scarlett, Press, 314

THE DEVELOPMENT OF SELF-CONCEPTIONS

and Crockett (1971) found that with increasing age, children use a greater number of constructs to describe their peers and that the proportion of egocentric and concrete descriptions declines while the proportion of nonegocentric and abstract descriptions increases. More recently, Bigner (1974) found similar developmental changes to occur between kindergarten and the eighth grade in children's descriptions of their siblings. In general, research on the development of person perception suggests that with increasing age, other people are viewed in a more interpersonal, complex and abstract manner (Peevers & Secord, 1973; Hill & Palmquist, Note 1). The results of these studies demonstrate the utility of applying a cognitive-structural approach to the issue of the development of person perception and suggest that such an approach may be advantageously applied to the development of other aspects of social cognition, such as self-conceptions. As a general statement about development, the orthogenetic principle suggests that an individual' s increasing ability to think abstractly not only results in the greater use of psychological and abstract constructs to describe others but also a correspondingly greater use of these types of constructs to describe the self. In addition, social psychologists consider the knowledge that an individual acquires about himself and about others to be the result of social interaction (Mead, 1934). For these two reasons, self-concept development may show a sequence of development that parallels the sequence found for the development of person perception. The purpose of this study is to extend the cognitive-structural perspective to the area of self-concept development. Based upon this orientation, it is hypothesized that young children primarily conceive of and describe themselves in terms of such concrete characteristics as their physical appearance and possessions, while adolescents conceive of themselves more abstractly and describe themselves in more psychological and interpersonal terms. In addition, data collected by Mullener and Laird (1971) suggest that between the ages of 12 and 29 years self-evaluations become increasingly more differentiated and less

315

global. A similar trend may also occur for self-concept development between childhood and adolescence and may be reflected in the adolescent's use of a greater variety of constructs to describe himself. Method Subjects Subjects were 136 males and 126 females drawn from four schools in a suburban, midwestern, university community. Approximately equal numbers of males and females from Grades 4, 6, 8, 10, and 12 completed the test instrument. The mean ages in years for students in each respective grade were 9.8, 11.8, 14.0, 15.9, and 17.9. The students were almost exclusively white and within an average and above average range of intelligence as indicated by academic performance and teacher evaluation. Almost all of the parents were in professional, entrepreneurial, or upper white-collar classes—Classes I and II according to Hollingshead's (Note 2) two-factor index of social position.

Procedure Students were administered the Twenty Statements Test (Burgental & Zelen, 1950; Kuhn & McPartland, 1954) in class groups. Students were given a test form with 20 spaces and were asked to write 20 different answers to the question "Who am I?"

Scoring System A 30-category scoring system developed by Gordon (1968) was used to classify each answer. Each answer was assigned to the one category that was judged to most accurately reflect the meaning of the response. The system was designed to capture the major varieties of self-representations and the categories are mutually exclusive and exhaustive. The 30 categories are listed in Table 1. Two undergraduates were trained in the use of the scoring system. Interjudge agreement was tested by having both coders independently score a random sample of 20 tests drawn from the original sample. These students' data were not part of the final analysis. Responses assigned to the same category by both coders were scored as agreed. Interjudge agreement was 85% (mean agreement per test, 17 out of 20 responses).

Results The data were summarized in terms of the percentage of students who used each category at least once. Since the sample size was large, and many chi-square tests were performed, only p values less than .001 were

316

RAYMOND MONTEMAYOR AND MARVIN EISEN

Table 1: Percentage of Subjects at Each Age Using Category at Least Once Age

10

12

14

16

18

73 35 10 4 0 28 12 59 0 0 16 39 34 80 14 65 28 36 62 22 57 23 8 0 37 76 42 23 7 15

38 30 8 2 4 18 29 37 4 0 21 34 19 31 24 80 40 30 82 24 46 17 26 15 44 91 65 24 20 10

48 25 11 13 5 25 28 54 3 2 13 38 26 45 24 45 24 28 75 14 49 28 45 17 48 86 81 28 20 6

72***

Age Name Racial or national heritage Religion Kinship role Occupational role Student role Political affiliation Social status Territoriality, citizenship Membership in actual interacting group Existential, individuating Membership in an abstract category Ideological and belief references Judgments, tastes, likes Intellectual concerns Artistic activities Other activities Possessions, resources Physical self, body image Sense of moral worth Sense of self-determination Sense of unity Sense of competence Interpersonal style Psychic style Judgments imputed to others Situational references Uncodable responses

45 18 50 5 7 37 4 67 0 4 48 57 0 2 4 69 36 23 63 53 87 4 5 0 36 42 27 23 9 19

n

53

50

55

65

Category Sex

41 31*"

15 10 57*** 44*** 72**

5 3 11***

57 54*** 52*** 39*** 31***

23 18 60 8*** 16***

26*

49*** 21***

36 93*** 72***

57" 10 8 39

Note. df = 4 in all cases. * Chi-square significant at .05 level. ** Chi-square significant at .01 level. *** Chi-square significant at .001 level.

considered reliable. With the age factor collapsed, chi-square tests for sex differences were computed for each category. No reliable differences were observed and the data for both sexes were combined for all subsequent analyses (N = 262). Statistically reliable age changes in self-conceptions were found for 15 of the 30 categories. Table 1 shows the percentage of students at each age using a particular category at least once. There were significant increases between childhood and adolescence in the percentage of subjects who used the following seven categories: occupational role (e.g., hoping to become a doctor, paper-boy), x2 (4) = 26.13; existential, individuating (e.g., me, I, myself), x 2 (4) = 38.24; ideological and belief references (e.g., a liberal, a pacifist), x2 (4) = 19.10; sense of self-determination (e.g., ambitious, a hardworker), x 2 (4) = 41.63; sense of unity (e.g., in harmony,

mixed-up), x 2 (4) = 20.38; interpersonal style (how I typically act, e.g., friendly, fair, shy),x 2 (4) = 48.44; and psychic style (how I typically think and feel, e.g., happy, calm), X2 (4) = 45.30. Adolescents were more likely than children to refer to themselves with terms that were future oriented, abstract, interpersonal, and psychological. There were significant decreases between childhood and adolescence in the percentage of subjects who used the following three categories: territorially, citizenship (e.g., an American, living on Oak Street, x 2 (4) = 27.95; possessions, resources (e.g., own a dog, have a bike), x 2 (4) = 33.72; and physical self, body image (e.g., 5' 10", fat),x 2 (4) = 48.13. These categories generally indicate a concrete description of one's address, material possessions, and physical self, such as height or weight. Finally, there were significant curvilinear

THE DEVELOPMENT OF SELF-CONCEPTIONS

age changes in the percentage of subjects who used the following five categories: sex (e.g., a girl, a guy), x 2 (4) = 19.62; name (e.g., Shirley), x2 (4) =2 44.89; kinship role (e.g., a son, a sister), x (4) = 18.53; membership in an abstract category (e.g., a person, a human, a speck in the universe), x 2 (4) = 69.41; and judgments, tastes, likes (e.g., hate school, like sports), x 2 (4) = 30.53. One's sex, name, and kinship status are concrete, objective aspects of the self, and the use of terms referring to these characteristics produced U-shaped developmental functions. The abstract designation of oneself as "a person" or "a human" reached a surprising peak at age 12, declined at age 14, and steadily increased from age 14 onward. Evaluation of one's activities, such as "like to play baseball" and "hate to clean my room," were classified judgments, tastes, likes. These responses contain both concrete and abstract aspects, and their use increased until age 14 and declined thereafter. The mean number of categories used at least once by students at each age was as follows: 8.6,9.0,9.1, 9.5, and 11.1; F(4, 257) = 3.86, ns. Although adolescents used more categories than children in describing themselves, the difference was not significant.

317

The concrete-abstract change is not a simple linear one, however, since additional findings suggest that curvilinear changes occur in the use of categories that could be considered either concrete or abstract. These changes primarily involve the use of concrete description by adolescents rather than the use of abstract descriptions by children. For example, many adolescents referred to concrete characteristics such as their sex and name when describing themselves, suggesting that this type of information has an important phenomenological meaning even to individuals who characteristically define themselves in more abstract terms. Alternatively, few young children described themselves in abstract terms with the exception of the many 12year-olds who referred to themselves as "a person" or "a human," abstract ideas which may reflect the initial appearance of advanced cognitive skills. The presentation of a few protocols may provide a more vivid picture of self-concept change in children and adolescents. The following examples were chosen to represent typical self-descriptions at different ages. The original spellings and emphases have been retained. The first set of responses are from a 9-year-old boy in the fourth grade. Note the Discussion concrete flavor of his self-descriptions and the almost exclusive use of terms referring The results of this study support the gen- to his sex, age, name, territory, likes, and eral hypothesis that with increasing age an physical self: individual's self-concept becomes more abstract and less concrete. The children in My name is Bruce C. I have brown eyes. I have brown this study primarily describe themselves in hair. I have brown eyebrows. I'm nine years old. I terms of concrete, objective categories such LOVE! Sports. I have seven people in my family. I great! eye site. I have lots! of friends. Hive on 1923 as their address, physical appearance, pos- have Pinecrest Dr. I'mgoingon lOin September. I'maboy. sessions, and play activities, while adoles- I have a uncle that is almost 7 feet tall. My school is cents used more abstract and subjective de- Pinecrest. My teacher is Mrs. V. I play Hockey! I'am scriptions such as personal beliefs, motiva- almost the smartest boy in the class. I LOVE! food. I tional and interpersonal characteristics. love freash air. I LOVE School. These overall findings are in general The next protocol is from a girl aged 1 Wi agreement with data from studies of the development of social cognitions (Hill & in the sixth grade. Although she frequently Palmquist, Note 1). The data from the refers to her likes, she also emphasizes her present investigation are also similar to data interpersonal and personality characterisobtained by Livesley and Bromley (1973) tics. who studied self-concept development My name is A. I'm a human being. I'm a girl. I'm a among English school children and who truthful person. I'm not pretty. I do so-so in my studies. used a different scoring system. I'm a very good cellist. I'm a very good pianist. I'm a

318

RAYMOND MONTEMAYOR AND MARVIN EISEN

lescents seem to infer from their own behaviors the existence of underlying abilities, motives and a personality style. As Inhelder and Kaget (1958) point out, adolescent thinking is a "second order system" in that the adolescent does not solve problems in The final example is from a 17-year-old terms of concrete givens, but uses those girl in the twelfth grade. Note the strong concrete facts to form hypotheses about an emphasis on interpersonal description, underlying reality. What appears to be the characteristic mood states, and the large self for the child is only the set of elements number of ideological and belief references. from which the adolescent infers a set of I am a human being. I am a girl. I am an individual. I personal beliefs and psychic style that don't know who 1 am. I am a Pisces. I am a moody uniquely characterize himself.

little bit tall for my age. I like several boys. Hike several girls. I'm old-fashioned. I play tennis. I am a very good swimmer. I try to be helpful. I'm always ready to be friends with anybody. Mostly I'm good, but I lose my temper. I'm not well-liked by some girls and boys. I don't know if I'm liked by boys or not.

person. I am an indecisive person. I am an ambitious person. I am a very curious person. I am not an individual. I am a loner. I am an American (God help me). I am a Democrat. I am a liberal person. I am a radical. lama conservative. I am a pseudoliberal. I am an atheist. I am not a classifiable person (i.e., I don't want to be).

One notes in these examples and in the protocols from our other subjects a developmental increase in the depth and vividness of self-conceptions. Children describe where they live, what they look like, and what they do. Their self-concept seems somewhat shallow and undifferentiated, both from other people and from their environment. Adolescents, however, describe themselves in terms of their beliefs and personality characteristics, qualities which are more essential and intrinsic to the self and which produce a picture of the self that is sharp and unique. Self-concept development is not an additive process. Adolescents do not simply add more complex and abstract ideas about themselves to their earlier, childish, concrete conceptions. In comparison to children, adolescents conceive of themselves quite differently; earlier notions either drop out or are integrated into a more complex picture. Self-conceptions appear to undergo a developmental transformation, perhaps based on the developing ability of the individual to draw inferences and form hypotheses about underlying characteristics. It is not uncommon, for example, for a young child to say that he likes to play baseball, football, hockey, and so on. An adolescent rarely responds in this way. A much more common response would be for him to say, "I am an athlete" or "I like athletics." Ado-

REFERENCE NOTES

1. Hill, J. P., & Palmquist, W. J. Social cognition and social relations in adolescence: A precursory view. Paper presented at the annual meeting of the Eastern Psychological Association, Philadelphia, April 1974. 2. Hollingshead, A. B. Two-factor index of social position. Unpublished manuscript, 1957. REFERENCES

Bigner, J. J. A Wernerian developmental analysis of children's descriptions of siblings. Child Development, 1974,45, 317-323. Bugental, J. F. T., & Zelen, S. L. Investigations into the "self-concept." I. The W-A-Y technique. Journal of Personality, 1950, 18, 483^98. Carlson, R. Stability and change in the adolescent's self-image. Child Development, 1965, 36, 659-666. Crockett, W. H. Cognitive complexity and impression formation. In B. A. Maher (Ed.), Progress in experimental personality research (Vol. 2). New York: Academic Press, 1965. Elder, O. H. Adolescent socialization and development. In E. F. Borgatta & W. W. Lambert (Eds.), Handbook of personality theory and research. Chicago: Rand McNally, 1968. Engel, M. The stability of the self-concept in adolescence. Journal of Abnormal and Social Psychology, 1959, 58, 211-215. Erikson, E. H. Identity: Youth and crisis. New York: Norton, 1968. Freud, A. Adolescence as a developmental disturbance. In G. Caplan & S. Lebovici (Eds.), Adolescence: Psvchosocial perspectives. New York: Basic Books, 1969. Gordon, C. Self-conceptions: Configurations of content. In G. Gordon & K. J. Gergen (Eds.), The self in social interaction (Vol. 1). New York: Wiley, 1968. Inhelder, B., & Piaget, J. The growth of logical thinking from childhood to adolescence. New York: Basic Books, 1958. Kokenes, B. Grade level differences in factors of selfesteem. Developmental Psychology, 1974, 10, 954958.

THE DEVELOPMENT OF SELF-CONCEPTIONS Kuhn, M. H.,& McPartland, T. S. An empirical investigation of self-attitudes. American Sociological Review, 1954, 19, 68-76. Livesley.W. J.,&Bromley, D. B.Personperceptionin childhood and adolescence. New York: Wiley, 1973. Mead, G. H. Mind, self, and society. Chicago: University of Chicago Press, 1934. Monge R. H. Developmental trends in factors of adolescent self-concept. Developmental Psychology, 1973, 8, 382-393. Mullener, N., & Laird, J. D. Some developmental changes in the organization of self-evaluations. Developmental Psychology, 1971,5, 233-236. Peevers, B. H., & Secord, P. F. Developmental changes in attribution of descriptive concepts toper-

sons. Journalof Personality and Social Psychology, 1973, 27, 120-128. Scarlett, H. H., Press, A. N., & Crockett, W. H. Children's descriptions of peers: A Wernerian developmental analysis. Child Development, 1971, 42, 439-453. Werner, H. The concept of development from a comparative and organismic point of view. In D. B. Harris (Ed.), The concept of development. Minneapolis: University of Minnesota Press, 1957. Werner, H. Comparative psychology of mental (few/opment. New York: Science Editions, 1961. (Received August 10, 1976)

Acknowledgment

In addition to the staff of regular consulting editors, other scholars are often asked to help in reviewing manuscripts submitted to the Journal. The Editor wishes to acknowledge with thanks the advice and consultation provided by the following persons who served recently as guest editors. Barbara Abrahams Wesley Allinsmith Millie Almy Teresa Amabile Barbara Anderson Steven R. Asher Roger Bakeman Alfred L. Baldwin Albert Bandura Robert A. Baron Thomas W. Bartsch Sandra L. Bern W. Keith Berg Marion S. Blank Jeanne H. Block Robert H. Bradley Lila G. Braine James Brown Robert B. Cairns Courtney B. Cazden Victor G. Cicirelli John C. Condry Philip S. Dale Carol S. Dweck

319

David Elkind Lucy R. Ferguson Alan M. Fink Mary Ann Fisher John H. Flavell Gregory T. Fouts Joel S. Freund Lynette Friedrich-Cofer Jean B. Gleason Roberta M. Golinkoff Chad Gordon John M. Gottman Gerald Gratch Stephen Greenspan Margaret A. Hagen Lynn A. Hasher John P. Hill Keith Holyoak Alice M. Isen Jerome Kagan Charles B. Keasey George A. Kellas Glenn Kleiman

Robert P. Klein Eric Klinger Nathan Kogan Lawrence Kohlberg Gisela Labouvie-Vief Gary Y. Larsen Eleanor E. Maccoby Gene R. Medinnus Carolyn Mervis Herbert L. Pick Michael Pressley Kenneth H. Rubin J. Philippe Rushton Sarah M. Ryan Herbert D. Saltzstein Earl S. Schaefer Robert L. Selman Robert S. Siegler Elizabeth L. Simpson Janet T. Spence Richard C. Teevan Cynthia E. Turnure Helen West