Adolescent identity exploration: a test of Erikson's theory of transitional crisis.

Kidwell, Jeannie S. ; Dunham, Richard M. ; Bacho, Roderick A. 等

According to Erikson, exploration is at the heart of the adolescent transition. In his words, exploration is the identity crisis, although crises are not necessarily acute or severe (Erikson, 1959). The remaking of personal identity, which is necessary to complete the transition, was said to be dependent on exploration. In a similar perspective, Grotevant (1987) has referred to exploration as the "work" of the identity process.

Erikson has characterized adolescent identity exploration as being accompanied by fluctuations in ego strength. Cognitive destructuring, generally, and the view of the self, in particular, was seen to result in reduced ego strength and impairment of coping. Moreover, a variety of symptoms were said to occur with the transition; these included subjective discomfort, confusion, mood swings, ego defenses, impulsivity, acting-out, and heightened physical and somatic complaints (Erikson, 1956, 1963, 1968).

Thus, for Erikson (1968), the self in transition is vulnerable: "Each stage becomes a crisis because incipient growth and awareness in a significant part function goes together with the shift in instinctual energy and yet causes specific vulnerability in that part" (Erikson, 1959, p. 56). Under such a challenge, the experience is one of a "... split of self images, a loss of center, and a dispersion" (Erikson, 1968). These symptoms and the experience of the self as "disrupted" have been described as the "... dark and negative side of identity formation," and they are viewed as vital to the identity process (Erikson, 1975).

The present research sought to find empirical evidence concerning Erikson's view that adolescent identity exploration is associated with reduced ego strength and the occurrence of psychological and physical symptoms. While much has been written in an attempt to understand the implications of the "crisis" of adolescence generally, the empirical research has not employed identity exploration as a key variable.

The general hypothesis of the present study can be stated as follows: If exploration is the identity crisis, then one would expect an adolescent who scores higher on measures of identity exploration also to show evidence of the symptoms as ascribed by Erikson.

METHOD

The sample consisted of 82 academically superior high school students (43 males; 39 females) between 14 and 17 years of age who had completed their junior year. It was comprised of two groups of students attending the Florida State University Summer Science and Mathematics Camp during the summers of 1988 (30 students) and 1989 (52 students). The minimum requirements for admission to the program included percentile ranks of 90 or more on standardized achievement tests, as well as sustained high academic performance.

The 1988 and 1989 groups did not differ in measures of socioeconomic status (SES) (Hollingshead, 1965). The SES of the families-of-origin for both groups was middle-to-upper-middle class. The mean SES index was 47.03, with a range of 31-62 for the 1988 sample and a mean of 48.96 with a range of 29-66 for the 1989 sample (Hollingshead Four Factor Index of Social Status, 1965).

Measures

The instruments used were the Minnesota Multiphasic Personality Inventory (MMPI) and the Ego-Identity Interview (Grotevant & Cooper, 1981).

Twenty-three of the commonly used scales of the MMPI were selected to assess symptomatology. There were the nine clinical scales and the 14 Frequently Scored Scales (see Graham, 1987). (Of the ten clinical profile scales, one - Mf, masculinity-femininity - was omitted from the analysis because of its differential norming for males and females.) Adolescent norms were used for scoring the clinical scales (Archer, 1987). Separate adolescent norms were not available for the Frequency Scored Scales.

The Ego-Identity Interview (Grotevant & Cooper, 1981) was based on Marcia's (1964) operationalization of Erikson's theory (1968) regarding adolescence as a time of increased exploration and commitment. The semi-structured interview was designed to gather evidence of exploration and commitment in six domains: occupation, religion, politics, friendships, dating, and sex roles. It examined both the depth and the breadth of exploration which influenced the values, beliefs, and goals of the adolescent. Depth was determined by the number of exploratory activities completed prior to making a commitment. Such activities might include reading about the topic, talking with others, and attending programs. The breadth of exploration was ascertained by the number of alternatives considered (e.g., how many different occupations or religious views considered). Questions also attempted to tap the extent of the "struggle" (e.g., disagreements or conflicts with parents or peers; doubts about beliefs or ideas).

Procedure

Each subject scheduled approximately two hours with the researchers. The paper-and-pencil measures were completed first, followed by the interview. Subjects were told that the purpose of the research was to examine personal development and family relations in adolescence.

Four student participants in the camp (two each in 1988 and 1989) were assigned to the research team as part of the program's aim to provide direct individual involvement in scientific research. They were trained over a two-week period to administer the instruments and to interview their peers in the camp. Their training included scoring and rating procedures. Each interviewer interviewed half of the subjects. All interviews were audiotaped. The students also were given training in the ethics of confidentiality. Science ethics in general was included as part of the formal instruction in the camp. No reports of abuse of the confidentiality standard were reported or observed.

The taped interviews were transcribed following the criteria offered in the Grotevant & Cooper (1981) manual which provides criteria from which to score exploration and commitment in each domain. After a training phase, each interviewer independently scored every interview. Raters scored each subject on four-point scales for exploration and for commitment in each of the six domains.

Exploration and commitment scores each were summed to obtain a total exploration score and a total commitment score. For both exploration and commitment, the possible range was from six to 24 (the sum of the ratings from one to four across the six domains). The mean exploration score for the combined samples was 14.6 (SD 2.8) with a range of 9-23. The mean commitment score was 16.9 (SD 2.5) with a range of 10-23. The scores on exploration were similar to those found by Grotevant & Cooper (1985) in a sample of high school seniors. They reported only the range of scores: 10-21 for males, 10-23 for females. These authors did not report scores on commitment.

The use of peers as raters is consistent with Archer and Waterman's (1983) call for the use of age-appropriate criteria for exploration and commitment: "The same statements made by a high school student and an adult would be interpreted quite differently as a function of their prior life-histories and the social expectations held for each. What may be seen as a relatively sophisticated view offered by the former could be considered naive if expressed by the latter." (p. 212)

Reliability

Reliabilities were calculated in a manner similar to that used by Grotevant and Cooper (1985). These were based on the number of agreements between two raters, divided by the total number of judgments in ratings of exploration and commitment across the six domains of identity. The mean reliability was .80 for exploration and .82 for commitment. Differences between raters were resolved by mutual consent between the two interviewers, with a senior researcher overseeing the process and attempting to facilitate their dialogue but not guiding their decisions. Two senior researchers also conducted random checks on the accuracy of the use of scoring procedures throughout the rating phase. The observed reliability of the exploration measure approximated that of Grotevant and Cooper (1985) who did not report a reliability for the commitment measure.

RESULTS

The 23 MMPI scales were correlated with exploration and with commitment. Only one of the scales correlated with commitment. In contrast, ten of the 23 scales were significantly correlated with exploration at the .05 level or better. These were Hs (Hypochondriasis), r = .30, p [less than] .01; Hy (Hysteria), r = .25, p [less than] .05; Pd (Psychopathic Deviate), r = .23,p [less than] .05; Pa (Paranoid), r = .28, p [less than] .01; Ma (Hypomania), r = .26, p [less than] .05; R (Repression), r = -.25, p [less than] .05; RE (Social Responsibility), r = -.25, p [less than] .05; MAS (Manifest Anxiety), r = .22, p [less than] .05; ES (Ego Strength), r = -.25, p [less than] .05; MAC (McAndrew Alcoholism), r = .30, p [less than] .01.

An exploratory factor analysis (principal components with Varimax rotation) was conducted, employing the ten significant scales. A one-factor solution was found on the basis of simplicity of structure, eigen values, and percentage of variance explained. It was labeled the Identity Exploration Crisis (IEC) factor. The symptoms isolated in the IEC factor are interpreted in terms of Erikson's description of the adolescent crisis. Broadly stated, the "crisis" symptoms, as formulated by Erikson and as captured in the IEC factor, include reduced Ego Strength (negative loading of the Ego Strength scale), impulsivity and acting-out (positive loadings of Psychopathic Deviate, Hypomania, and McAndrew Alcoholism), heightened physical and somatic complaints (positive loading of Hypochondriasis and Hysteria). The seven variables loading on IEC and their factor loadings are shown in Table 1.

The associations among the factor scores are all in the hypothesized direction, with only Ego Strength showing a negative association, as predicted.

DISCUSSION

The present findings support Erikson's concept that exploration is at the heart of the identity crisis. Consistent with Erikson's assertions, the IEC factor isolated in this study suggests that identity exploration is associated with a decline in ego strength, and with symptoms related to the use of ego defenses. Three of the scales which load on IEC (Psychopathic Deviate, Hypomania, and McAndrew Alcoholism) capture primarily impulsivity, acting out, and imperturbability, while Hypochondriasis, Hysteria, and Paranoia represent repression, projection, denial, and reaction formation. The decline in ego strength implies a reduced capacity for coping with problems and stress.

For a more fine-grained description of the signs and symptoms of adolescent identity exploration, as found in this study, we can draw on the work of Graham (1987). Using common interpretations of the MMPI scales comprising the IEC factor, we can say that adolescents who are actively involved in the exploration process are more likely than their low-exploring peers to manifest inner confusion, agitation, dissatisfaction, unhappiness, periodic episodes of depression, a vacillation between poor self-concept and grandiosity, and disturbed thinking. They will tend to lack insight into their own motivations and to feel misunderstood by authorities. They will fret and worry, feel confused, look at the negative side, and not understand themselves well. They may be more likely to experience increased physical symptoms. Table 1

Factor Loadings of MMPI Scales Associated with Adolescent Identity Exploration

Variable Factor Loading

PD - Psychopathic Deviate .86 HS - Hypochondriasis .78 HY - Hysteria .72 PA - Paranoia .66 MA - Hypomania .53 MAC - MacAndrew Alcoholism .23 ES - Ego Strength -.31

Percent of Variance = 39.1

Eigenvalue = 2.7

In interactions with others, adolescents who are exploring may fluctuate between an ongoing, energetic, impulsive approach and a bored, withdrawn, seclusive response. They may tend toward poor social judgment and may be rebellious and hostile toward parents and other authority figures, whom they tend to blame for their own problems. They may be excessively conforming to peer pressures (Graham, 1987).

The above descriptions lend credence to the assertions by Grotevant (1987) and by Roll (1980) that the process of exploration is both intra-psychic and interpersonal. (For a discussion of the interplay of societal, interpersonal, and intrapsychic factors in the adolescent passage, see Dunham, Kidwell, & Wilson, 1986.)

The present findings, furthermore, support the writings of others that the formation of symptoms may occur when the ego is undergoing a destructuring process, i.e., during periods of active self exploration (Josselyn, 1973; Cote & Levine, 1987; Chandler, Boyes, & Ball, 1990).

It is interesting to contemplate the joint implications of the Hypochondriasis and Ego Strength measures. As exploration increases, there is a tendency for bodily concerns to be expressed and for ego strength to decline. Hypochondriasis and Ego Strength both relate in a relatively direct way to the perception of self-adequacy. Thus, if this work is replicated, it may be possible to see cognitive destructuring as affecting both the psychological and the physical self.

It also should be pointed out that the MMPI scores in this study were not indicative of symptoms at levels comparable to clinical patients. The mean clinical profile of this sample was consistent with that of "normal" adolescents, as described by Archer (1984). These students were highly successful, carefully selected high schoolers, who by all accounts would be defined as "good, healthy kids" by their parents and by their teachers. Nevertheless, the findings suggest that there may be certain symptom patterns in even the most "healthy" adolescents, especially when they are actively exploring their identities.

Perhaps, the findings also are consistent with Erikson's (1968) argument that the period of active exploration is likely to be pronounced in the gifted. A similar pattern of findings may not be revealed in a less select group of comparable age. Perhaps the "crisis" will surface at a later age, or perhaps the profiles will differ. Such questions remain to be addressed in future research.

Additionally, these adolescents were involved in a new and unfamiliar environment, being away from their families and living on a college campus for a six-week period. This may imply that the exploration process is encouraged by virtue of exposure to new and differing points of view (Grotevant, 1987).

The lack of relationship between commitment and the MMPI scales contrasts with the findings that exploration is strongly associated with a pattern of MMPI scales. The commitment score can be seen as a control, of sorts. It provides some assurance that the correlations with exploration are meaningful; the MMPI does not automatically assign meaning to a measure. The absence of a pattern of psychological "disturbance" in those who are "committed" may also reflect a psychological "relaxation," to the extent that commitment implies an internal organization or structure, whereby the individual feels a sense of purpose, meaning, and direction (Waterman, 1984). In support of this view, Cote and Levine (1987), in their interpretation of Erikson's work, have suggested that when the ego identity is intact, i.e., "structured," the personality is more unified; the elements of the self are cohesive, and the personality is more integrated with the social world; that is, the self-concept is embedded within and validated by the successful enactment of social roles.

A replication of this study is needed. There is, to our knowledge, no direct counterpart. The use of a continuous measure of exploration in relationship to measures of predicted symptoms, together with factor analysis or other multivariate techniques to isolate meaningful clusters is new to the identity research literature. With additional studies using this method on different samples, perhaps the symptomatology associated with exploration will be found to be meaningfully patterned. Future studies may draw on an array of populations, circumstances, developmental stages, and be done with alternative measures and competing hypotheses.

The present findings may prove to be useful in broadening the concept of ego identity formation beyond the more narrow stage-specific conceptualization. One might hypothesize, for example, that at any crucial developmental transition throughout the life span, the individual may experience heightened symptoms of varying nature and degree during which the "self" is being reconstructed. It is a promising lead toward additional research.

Finally, the results of this study present a picture of the "dark and negative side" (Erikson, 1975) of identity formation where the self is experienced as vulnerable, confused, and conflicted. The present authors suggest that these findings may provide a synthesis of the adolescent "storm and stress" dialectic (see Coleman, 1978 for a discussion.) As suggested by Berzonsky (1982), what is needed is to begin to isolate the conditions which contribute to the formation of stress symptoms.

Erikson (1968) has stressed the need for supportive interactions with significant others in order to provide strength during phases of ego destructuring. If the results of the present study are sustained and elaborated, they will be helpful in guiding adolescents, their parents, and others in sophisticated patterns of support in sustaining positive gains in identity.

REFERENCES

Archer, R. P. (1984). Use of the MMPI with adolescents: A review of the salient issues. Clinical Psychology Review, 4, 241-251.

Archer, R. P. (1987). Using the MMPI with adolescents. Hillsdale, NJ: Lawrence Erlbaum.

Archer, S. L., & Waterman, A. S. (1983). Identity in early adolescence: A developmental perspective. Journal of Early Adolescence, 3, 203-214.

Berzonsky, M.D. (1982). Inter- and intraindividual differences in adolescent storm and stress: A life-span developmental view. Journal of Early Adolescence, 2, 211-217.

Chandler, M., Boyes, M., & Ball, L. (1990). Relativism and stations of epistimic doubt. Journal of Experimental Child Psychology, 50, 370-395.

Cote, J. E., & Levine, C. (1987). A formulation of Erikson's theory of ego identity formation. Developmental Review, 7, 273-325.

Dunham, R. M., Kidwell, J. S., & Wilson, S. M. (1986). Rites of passage at adolescence: A ritual process paradigm. Journal of Adolescent Research, 1, 139-153.

Erikson, E. H. (1956). The problem of ego identity. Journal of the American Psychoanalytic Association, 4, 56-121.

Erikson, E. H. (1959). Late adolescence. In D.H. Funkenstein (Ed.), The student and mental health, Cambridge, MA: Riverside Press.

Erikson, E. H. (1963). Childhood and society. New York: Norton.

Erikson, E. H. (1968). Identity: Youth and crisis. New York: Norton.

Erikson, E. H. (1975). Life history and the historical moment. New York: Norton.

Graham, J. R. (1987). The MMPI: A practical guide, (2nd ed.). New York: Oxford University Press.

Grotevant, H. D. (1987). Toward a process model of identity formation. Journal of Adolescent Research, 2, 203-222.

Grotevant, H. D., & Cooper, C. R. (1981). Assessing adolescent identity in the areas of occupation, religion, politics, friendships, dating, and sex roles: Manual for administration and coding of the interview. JSAS Catalog of Selected Documents in Psychology.

Grotevant, H. D., & Cooper, C. R. (1985). Patterns of interaction in family relationships and the development of identity exploration in adolescence. Child Development, 56, 415-428.

Hollingshead, A. B. (1965). Index of social position. New Haven, CT: Yale University Press.

Josselson, R. L. (1973). Psychodynamic aspects of identity formation in college women. Journal of Youth and Adolescence, 2, 3-52.

Marcia, J. E. (1964). Determination and construct validity of ego identity status. Unpublished doctoral dissertation. Ohio State University, Columbus, OH.

Roll, E. J. (1980). Psychologists' conflicts about the inevitability of conflict during adolescence: An attempt at reconciliation. Adolescence, 15, 661-670.

Waterman, A. S. (1984). Identity formation: Discovery or creation? Journal of Early Adolescence, 4, 329-341.

Richard M. Dunham, Ph.D., Professor, Department of Psychology, Florida State University.

Roderick A. Bacho, M.A., Department of Psychology, Florida State University.

Ellen Pastorino, Ph.D., Assistant Professor, Department of Psychology, Gainesville College.

Pedro R. Portes, Ph.D., Associate Professor, Department of Educational Psychology, University of Louisville.