Pergamon
Perron. mdkk~. Dif/. Vol. 21. No. 6. pp. 1021-1027, 1996
Copynght ~CJ1996 Elsevier Science Ltd
Printed in Great Britain. All rights reserved
SO191-8869(96)00128-6 0191.8869!96 $lS.00+0.00
THE IMPORTANCE OF ADOLESCENTS’ SELVES:
DESCRIPTION, TYPOLOGY AND CONTEXT
Petr Macek’ and Lida Osecka’
‘Institute of Psychology, Masaryk University, A.Novaka I, 660 88 Bmo, and ‘Institute of Psychology,
Academy of Sciences, Veveri 97. 602 00 Bmo, Czech Republic
(Recked 14 October 1995; receiaed,for publicurion 26 June I996)
Summary-Real self and ideal self are seen as important determinants of self-satisfaction or self-acceptance.
Specific importance can be ascribed to other self-representations such as the unwanted self, and the self
according to significant others. Empirical study was oriented towards the description of subjective importance
of the self-modalities (total sample 745 subjects, a mean age 15.8 years). Results indicate a stability
of high importance self according to parents, and of the importance of the actual, ideal, and unwanted self
of adolescents. Three empirical types were identified: External anchored, less autonomous self-concept
(Type I), autonomously anchored, parents influenced self-concept (Type 2) and autonomously anchored.
peers influenced self-concept (Type 3). Copyright ,c 1996 Elsevier Science Ltd.
INTRODUCTION
Identity formation is considered to be the most important developmental task of adolescence
(Erikson, 1968; Petersen & Leffert, 1992). Within this context, self-definition, self-structuring and
self-understanding are central and necessary processes (Berzonsky, 1990).
The dynamics of the organization of the adolescent’s self-system may be understood in terms of
the ongoing processes of differentiation and integration, which are reflected in the various forms of
perceived selves and inter-relationships among them (Oosterwegel, 1992). Besides the analysis of
actual contents of these perceived selves, the knowledge of their personal relevance and their
subjective importance is the key for the understanding of adolescent self-development (Marsh, 1986;
Pelham & Swann, 1989).
Theoretically, we can consider many perceived selves: intraindividual and interindividual (Oosterwegel,
1992), real and normative selves, actual and possible selves (Markus & Nurius, 1986)
former, present, and future selves, selves according to significant others, etc. Some of them can be
very important for self-regulation and self-worth of adolescents, whilst some actual or possible
perceived selves can be of a small or very specific importance for an individual.
During a period of adolescence, we can assume changes in subjective importance of perceived
selves as well as many interindividual differences. However, traditionally major attention is focused
upon real (actual) self and ideal self (Assor & Tzelgov, 1987; Glick & Ziegler, 1985). Many
researchers a priori suppose that these two intrapersonal perceived selves are the most important
for adolescents. Discrepancies between the real and ideal self are often interpreted as an indicator
of cognitive development of adolescents or as an indicator of their self-acceptance.
In spite of this, the other perceived selves can be also of similar importance: the unwanted or
undesired self (Macek, 1991; Ogilvie, 1987) and many interpersonal perceived selves according to
significant others, i.e. parents, friends, adult authorities, etc. (see Higgins, 1987; Oosterwegel, 1992;
Smollar & Youniss, 1985; Spiel, 1992). Our recent pilot study (Macek, Osecka & Blatnjr, 1992)
revealed, that especially the perceived self according to parents, preserves its high importance stably
during the period of adolescence. The importance of other perceived selves, such as the real self,
ideal self, self according to friends, etc., was more dependent on gender and the age of adolescents.
However, the importance of perceived selves is only one cognitive aspect of a process of self
definition. Another aspect of the self-system (affective components, cognitive structure and multidimensionality
of the self-concept) as well as other cognitive and psychosocial characteristics take
part in the process of adolescent identity formation (Berzonsky, 1990; Bosma, 1991; Kroger, 1989;
Nurmi, 1991). We have found that the importance of perceived selves is more closely related to the
1021
1022 Petr Macek and Lida Oseckii
different cognitive identity variables (cognitive styles, future expectation, coping strategies) than to
emotional (affective) aspects of the self-system (Macek & Oseck& 1992).
This study represents a further analysis of this matter. The first aim is to specify gender and age
differences in the importance of perceived selves in the age of mid-adolescence (14-17 years).
However, we can expect high interindividual variability here. Therefore, the second aim of our
study is to create an empirical typology of adolescents according to the importance of perceived
selves as a criterion. Such types (or subgroups) should demonstrate not only specific configurations
of important perceived selves, but also different relationships of these configurations to other
psychosocial characteristics of adolescents, which participate in a process of their identity formation.
METHOD
For purposes of this study, three adolescent samples were included. Sample I included 274
participants (117 girls, 157 boys) at two age levels: 14 and 16. Sample II included 171 participants
(111 girls, 60 boys) between ages of 15 and 19. Sample III included 300 participants (149 girls, 151
boys) at age 17. The data were collected during February 1992 (Sample I), during December 1992
(Sample III) and during May 1993 (Sample II).
The total sample consisted of 745 adolescents (377 girls, 368 boys), with a mean age of 15.8. Two
age groups were defined as follows: younger adolescents (N = 290, mean age 14.4) and older
adolescents (N = 455, mean age 16.6). Finally, Ss were divided into four groups: younger girls
(N = 134), younger boys (N = 156), older girls (N = 243) and older boys (N = 212).
The questionnaire measures
Our Self-Modalities Importance Measure (SMI) is based on a paired comparison of importance
of nine perceived selves: the actual, real self (statement: “How I am”); the ideal self (“How I would
like to be”); the unwanted self (“How I would not like to be”); the self according to parents (“How
I think that my parents see me”); the self according to friends (“How I think that my friends see
me”); the self according to adult authorities (“How I think that my favorite teacher, leader, etc.,
sees me”); the ideal according to parents (“How I think that my parents would like to see me”); the
ideal according to friends (“How I think that my friends would like to see me”); and the ideal
according to adult authorities (“How I think that my favorite teacher, leader, etc. would like to see
me”). These nine self-modalities are presented in the form of their 36 possible paired combinations.
(Example: “Decide, which is or more importance for you to know: 1. How am I? or 2. How would
I like to be?“).
The Identity Style Inventory (ISI) (Berzonsky, 1989, 1992) includes 39 items that are rated on a
Likert scale from 1 (not at all like me) to 5 (very much like me). Four separate scores are yielded
for: (a) Information oriented style (active seeking out and evaluation of an individual’s relevant
information before making a decision); (b) Normative oriented style (the orientation that is more
concerned with conformity to the normative standards and prescriptions of significant others); (c)
Avoidant/diffuse oriented style (procrastination and defensive avoidance of problems, situational
rewards dictate a course of action); and (4) Commitment (personal involvement in the beliefs,
norms, and values) (Berzonsky, 1989).
The Self-Esteem Scale (SES) and Life-Satisfaction Scale (LSS) are two factors from Bernse
Questionnaire of Adolescent’s Subjective Well-Being (namely factors self-value and positive attitude
toward life-see Grob, Liithi, Kaiser, Flammer, Mackinnon & Wearing, 1991). SES includes five
items, LSS includes eight items. Four points on the Likert Scale were used (from “totally false” to
“very true”). Total scores of self-esteem and life-satisfaction were calculated and used for the next
analysis.
The Daily Hassles Inventory (DHI), Coping Strategies Inventory (CSI), and Future Expectations
Inventory (FEI) were constructed for the purpose of the Euronet Pilot Study of Adolescents
(Flammer, Grob & Nurmi, 1991).
DHI includes 13 items, expressing common daily problems. Total score of daily hassles is
interpreted.
CSI contains 17 common reactions on stress and three summarized scores were used for the next
Importance of adolescents’ selves 1023
analysis: rational (active) coping, emotional (passive) coping, and interpersonal-family oriented
coping.
FE1 includes 16 possible future goals of adolescents. In respect to the aims of this study, five more
general domains of future expectations are used: the goal oriented to social prestige; the future
responsibility to others; the future leisure time; the future family life; and the future orientation to
basic life values (health, profession, etc.). All items were assessed on four points of the Likert scale.
SMI, ISI, SES, and LSS were presented to all participants. DHI, CSI, FE1 were presented to
Samples I and II only.
RESULTS
1. The importance of perceived selves
1.1. The total sample. Table 1 shows the means of the importance of self-modalities for the total
sample as well as for subgroups according to the gender and age criteria.
Within the total sample of adolescents, the most important of the perceived selves (self-modality)
is the self according to parents. It ranks first on the scale of important with relative distance from
other perceived selves. The real (actual) self, ideal self, self according to peers, and unwanted self
are assessed at a similar level of importance. Evidently, they are more important than remaining
perceived selves. Also, the ideal according to parents was reported as a relatively important self
modality of the adolescent’s self-system. The ideal according to peers and the actual self according
to authorities are less important. However, these two perceived selves are ranked far from the least
important self-modality-the ideal according to authorities.
1.2. The age and gender differences. Concerning sex and age differences, multivariate analysis
(MANOVA) and univariate analysis (ANOVA) of variance 2 x 2 (age subgroups x gender) was used
to analyse the importance of perceived selves.
Both multivariate effects are significant, that is, sex and age of adolescents influence the subjective
importance of perceived self modalities. No significant interactions of age x gender were found.
In regard to the real self, the age and gender differences were found to be significant. Real self
was more important for girls than boys and more important for older than younger adolescents.
However, it can be seen, that the importance of real self increases with age more rapidly for girls
than boys.
No significant age and gender differences are found concerning the importance of ideal self.
The importance of unwanted self is different in respect to age levels as well as in respect to girls
or boys. Especially for older adolescents unwanted self is more important than for younger ones.
Also, girls report higher importance of unwanted self than boys.
In respect to importance of self according to parents, a significant difference was found for age
levels only. This self-modality was more important for younger than for older adolescents.
Table I. Importance of self modalities: means for total sample groups of girls, boys, younger and older adolescents
Total Girls Boys Younger Older
Gender
d.f. = 1705
Age group
d.f. = 1705
Actual (real) self
Ideal self
Unwanted self
Self according to parents
Self according to peers
Self accordmg to authorities
Ideal according to parents
Ideal according to peers
Ideal according to authorities
Multivariate effects
4.74 4.97 4.52 4.48 4.91
4.72 4.77 4.67 4.73 4.72
4.67 4.92 4.41 4.25 4.93
5.32 5.46 5.17 5.46 5.23
4.67 4.91 4.43 4.41 4.83
2.91 2.85 2.97 3.06 2.82
3.87 3.65 4.10 4.18 3.61
3.01 2.89 3.14 2.83 3.12
1.57 1.30 I.87 I.77 I.45
F = 7.41 F= 6.21
P = 0.007 P = 0.013
F= 7.31 F = 12.62
F = 5.99
P = 0.015
F = 4.74
P = 0.03
F = 9.79
P = 0.002
F = 10.3
P = 0.001
F = 19.0
P = 0.000
F = 4.23
P < 0.001
F = 9.82
P = 0.002
F = 4.78
P < 0.029
F = 4.06
P = 0.044
F = 4.22
P < 0.001
Anova 2 x 2 (sex. age group). Only significant results are mentioned.
1024 Petr Macek and Lida Osecka
The importance of self according to friends depends on both independent variables. This selfmodality
is more important for girls than for boys and more important for older adolescents than
for younger ones. The effect of gender is higher than the effect of age.
Age and gender differences were not found in respect to the importance of self according to
authorities.
On the other hand, gender and age differences were significant in regard to the importance of
ideal self according to parents: this normative self is more important for boys than for girls and
more important for younger adolescents than for the older ones.
The ideal according to friends is less important for younger adolescents than for the older ones.
The effect of gender was not found.
Highly significant gender differences are typical for the ideal according to authorities-girls
reported its lower importance than boys. Also, the age differences were significant: ideal according
to the authorities was more important for younger adolescents than for the older ones.
2. The empirical typology
To describe specific groups (empirical types) with different profiles of important perceived selves,
solutions for two, three, four, and five clusters were tested (Quick Cluster, SPSS/PC+). The goal
was to find such clusters (subgroups), which would be different in respect to the subjective importance
of perceived selves. Moreover, clusters would have to include a sufficient number of Ss in respect to
the next statistical analyses.
Concerning these criteria, four clusters were selected. According to the Wilks criterion, the
combined dependent variables were significantly influenced by the independent variable ‘type’
(cluster), F(27,688) = 148.00, P < 0.001. Regarding the number of Ss, three clusters were comparable
(209, 267, and 219 Ss), the fourth included nine Ss only. That is why we describe, as
empirical types, the first three clusters only.
2.1. The type differences. As indicated in Table 2, significant type differences were found on all
nine dependent variables.
A chi-squared test revealed a significant association between types and age, x2(2) = 7.35, P < 0.05
and also between gender and types, x2(2) = 13.7, P < 0.005.
In respect to other psychosocial characteristics of identity process formation, significant univariate
effects were found (more details in Table 3) for:
l informational, normative, diffuse identity style and commitments;
l global level of life satisfaction;
l global sum of daily hassles;
l rational, emotional and on family oriented coping;
l importance of future leisure activities, future interpersonal responsibility, and future orientation
on general values (health, profession).
No significant differences between groups were found in respect to global self-esteem and future
goals orientation toward family life and social prestige.
Table 2. Means comparisons by types according to the importance of self-modalities
Type
I 2 3 F-test
Importance of.
Actual (real) self
Ideal self
Unwanted self
Self according to parents
Self according to peers
Self according to authorities
Ideal according to parents
Ideal according to peers
Ideal according to authorities
3.55ab
2.84ac
2.6Oab
6.16ab
5.27ab
4.29ab
4.70a
3.68a
2.79ab
5.38a
5.77ab
5.78a
5.68ac
3.23ac
2.38b
4.45b
1.93ab
1.29bc
5.31b 70.26
5.37bc 242.87
5.41b 211.08
4.29bc 80.35
6.12bc 255.19
2.32a 109.02
2.46ab 130.90
3.78b 105.38
0.75ac 104.03
Now: Within rows, means sharing a common single-letter differ significantly by at least P < 0.05 (Tukey-HSD
procedure): degrees of freedom for all unvariate F-test = 2,691. All F-tests are significant at least on P < 0.001
level.
Importance of adolescents’ selves 1025
Table 3. Type differences (means comparisons) in respect to different psychosocial characteristics (dependent variables)
Dependent
variables 1
TYPO
2 3 F-test P
Identity styles:
(a) I”formatlo”al 3.28a 3.45a 3.40 4.84
(b) Normative 3.40 3.44b 3.32b 3.07
(c) Diffusion 2.93~ 2.88b 3.06bc 6.43
Commitments 3.37ab 3.56ab 3.32b 9.36
Future goals orlented on:
(a) Social prestige 2.60 2.62 2.59
(b) Social responsibility 3.39a 3.40b 3.18ab I I.08
(c) Family life 3.27 3.36 3.42
(d) Pleasure leisure time 3.01 3.02 3.17 2.91
(e) Basic values 3.9la 3.88b 3.78ab 8.80
Coping strategies:
(a) Rational. active 2.69a 2.85a 2.76 3.46
(b) Emotional, passive l.9la I.92b 2.10ab 5.62
(c) On family oriented 2.25~1 2.3lb I.88ab 10.15
Daily hassles 1.94 I.89 2.00 2.73
Life satisfaction 2 88 2.94b 2.8lb 5.26
Self-esteem 2.80 2.85 2.82
Now Within rows, means sharing a common single-letter differ significantly by at least P < 0.05 (Tukey-HSD procedure).
0.008
0.05
0.002
0.0001
0.0000
0.055
0.0002
0.03
0.004
0.001
0.06
0.005
NS
To summarize, the obtained adolescent types (groups) are different in relevant self-variables
(importance of perceived selves) as well as in other psychosocial characteristics, which play an
important role in the process of identity formation.
2.2. The description of types.
Type 1: External Anchored, Less Autonomous Self-concept
These adolescents were the youngest of all (mean age 15.6 years, 88 girls and 119 boys). In
contrast to others, they evaluated the autonomous modalities of self (real, ideal, and unwanted self)
as less important. On the other hand, the self according to parents was found to be the most
important. Similarly, the other interpersonal anchored modalities of self (self according to peers,
self according to authorities) obtained a high importance rank. Also, some normative, interpersonal
anchored selves (the ideal according to parents, the ideal according to peers), were presented as
relatively important.
Concerning identity styles, normative orientation was found to be the most characteristic trait of
this group. Compared to other types, these adolescents reported lower score of information orientation
and rational coping strategies.
Type 2: Autonomously Anchored, Parents Influenced Self-concept
Concerning age, the second type is similar to the first. However, there are more girls than boys
here (143 girls, 120 boys, mean age 15.7 years).
The ideal self, unwanted self, real self, and also self according to parents were reported as highly
important. On the other hand, other interpersonal anchored self modalities (self according to peers,
self according to authorities) were found to be less important.
Regarding the identity formation, this group, more than others, preferred the information and
normative style and their level of commitment was also the highest. Moreover, they presented the
highest level of global-life satisfaction and the use of rational coping strategy more frequently than
others.
Type 3: Autonomously Anchored, Peers Influenced Self-concept
More girls (132) than boys (87) entered the third type. These adolescents are older than others
(mean age 16.2 years).
In contrast to previous types, these Ss reported the highest importance of self according to peers.
1026 Petr Macek and Lida Osecka
Autonomous selves were also highly important, self according to parents was found to be less
important. Also, normative selves were described as less important (with the exception of ideal
according to peers).
Compared to others, this groups presented higher scores of diffuse identity style, lower orientation
on social responsibility and basic values in the future. Family oriented coping was found less typical,
but the emotionally passive coping was more typical for these adolescents. Moreover, relatively
lower life satisfaction and more daily hassles were found to be their characteristic.
DISCUSSION
The data reported here may have several implications for research of self-system and for understanding
self-development in adolescence.
First, the subjective importance of perceived selves (self-modalities) is an important variable that
is necessary to include within the design of self-system research. Although some studies are relevant
(Markus & Nurius, 1986; Marsh, 1986; Oosterwegel, 1992; Pelham & Swann, 1989; Van der Werff,
1990), studies of self-concept rarely consider it.
Second, the importance of perceived selves is related to the age and gender of adolescents, as well
as to their personality and psychosocial context.
Concerning age and gender differences, we have found main effects only, but no interactions.
Therefore, we can describe some global developmental trends of importance of perceived selves.
Significant age differences were found for the importance of seven presented modalities of self.
Only the ideal self and self according to authorities do not change in respect to their importance.
The importance of the self according to parents decreases. However, the level of importance is
highest compared to the other ones.
The importance of the real, unwanted and also self according to friends increases during midadolescence.
For a group of older adolescents, all interpersonal normative (ideal) selves are apparently
less important than actual interpersonal selves.
Regarding the gender differences, two autonomous selves (real self, unwanted self) and self
according to peers are more important for girls than boys. Two normative selves, ideal according
to parents and ideal according to authorities, are more important for boys than girls. It seems
surprising that during mid-adolescence the importance of normative selves is higher for boys than
for girls. On the other hand, this result also means that the girls assign more importance to
intrapersonal and actual interpersonal selves.
Third, our results may contribute to the discussion about the role of parents and peers in the
process of identity formation. Great importance of the self according to parents for most adolescents
was confirmed. Some recent studies of adolescent self-concept show the meaning of relationships to
parents in various contexts (Spiel, 1992; Youniss & Smollar, 1990). Formerly presented opinion,
that importance of parents decreases during adolescence and importance of peers increases, should
be carefully revised. Parents keep a status of the most important persons, although the reasons for
this importance changes. Volpe (198 1) found that mid- and late-adolescents conceptualized their
relations to parents as more unilateral in form than did pre-adolescents or young adults. As
Oosterwegel (1992) has mentioned, relations to parents and peers are of a different meaning. The
parents are perceived in their role of setting standards that have to be met, the peers are sought for
empathic understanding.
The presented empirical typology offers three specific patterns regarding this matter. Adolescents,
who assign higher importance to peers and lower importance to parents, create a special subgroup
(Type 3). They also report other different characteristics in context of their identity formation.
Compared to others, they are more diffusely oriented, with a higher level of daily hassles. They are
more emotional, passive and less satisfied in life, and much less oriented on help from parents than
others.
The importance of self according to parents is encountered in two different configurations. On
the one hand, if it is univocally preferred over other selves, it can be seen as an indicator of higher
interpersonal dependence (and probably also lower stadium of identity formation-Type 1). On the
other hand, if both self according to parents and autonomous selves are equally important, it is no
indicator of proper adaptation and non-conflict in interpersonal relations.
Importance of adolescents’ selves 1027
REFERENCES
Assor, A. & Tzelgov, J. (I 987). Self-ideal discrepances as indicators of self-derogation and enhancement: Formalization of
theoretical claims and a method for assessment. Journal ofPersonality Assessment, 5/(4), 532-544.
Berzonsky, M. D. (1989). The self as a theorist: Individual differences in identity formation, International Journal q/
Personalir), Construct Psychology, 2, 363-376.
Berzonsky, M. D. (1990). Self-construction over the life span: A process perspective on identity formation. Adcances in
Personal Construct Psycholog!. I, 155-186.
Berzonsky, M. D. (1992). Social cognitive aspect of identity style. Journal of Adolescents Research, 7(2), 14&l 55.
Bosma. H. A. (1991). Cognitice comple.uiry and identity formation in adolescence. Presentation at the symposium: Process
and Context in Identity Formation, XIth Biennial Meetings of the ISSBD, 337 July, 1991, Minneapolis,
Erikson, E. H. (1968). Identify, yourh, and crisis. London: Faber & Faber.
Flammer, A., Grob, A. & Nurmi, J.-E. (1991). Euronet questionnaire. Bern: University of Berne.
Click. M. & Ziegler. E. (1985). Self-image: A cognitive developmental approach. In Leahy, R. L. (ed.) The decelopment of
the self(pp. l-53). New York: Academic Press.
Grob, A., Ltithi, R.. Kaiser, F. G.. Flammer, A.. Mackinnon, A. & Wearing, A. J. (1991). Berner fragebogen zum
wohlbefinden jugendlicher (BFW) (Bernse questionnaire of adolescent well-being). Diagnostica, 37, 6&75.
Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect, Psychology Review. 94, 319-340.
Kroger, J. (1989). Identit? in adolescence: The balance bettveen selfand other. Routledge, London.
Macek, P. (1991). Prispevek ke konceptualizaci obsahu sebepojeti. (Contribution concerning conceptualization of selfconcept).
Shornik praci FF BU. 125, Brno.
Macek. P. & Osecka, L. (1992). Identity, self and coping strategies of Czech adolescents as possible predictors of their Lvellheing.
Paper and poster presented at The Johann Jacobs Foundation’s Conference on ‘Youth in the Year 2000:
Psychological issues and interventions’. Marbach Castle, Germany, November 5-7.
Macek. P.. OseckB, L. & Blatny, M. (1992). Semantic structure of adolescents’ actual, ideal, and unwanted self. Studia
Minora Facubatis Philosophicae Unioersilatis Brunensis, I27, I I III 2 I, Bmo.
Markus, H. & Nurius, P. (1986). Possible selves. American Psychologist, 41, 954-969.
Marsh. W. H. (1986). Global self-esteem: Its relation to specific facets of self-concepts and their importance. Journal Q/
Personalit!, and Social Psychology, 51, 12241236.
Nurmi. J.-E. (1991). How do adolescents see their future’? A review of the development of future orientation and planning.
Decelopmental Rrriebr, II, l-59.
Ogilvie, D. M. (1987). The undesired self: A neglected variable in personality research. Journal of Personality and Social
Psychology. 52. 379-385.
Oosterwegel, A. (1992). The organization ofthe self-system: Derelopmenial changes in childhood and adolescence. Amsterdam:
Universiteit van Amsterdam.
Pelham. B. W. & Swann. W. B. (1989). From self-conceptions to self-worth: On the sources and structure of global selfesteem.
Journal of Personality and Social Psychology, 57, 672-680.
Petersen, A. C. & Leffert. N. (1992). What is special about adolescence.7 Paper presented at a conference ‘Youth in the Year
2000: Psychosocial Issue and Interventions’. Johann Jacobs Research Centre, Marbach Castle, Germany.
Smollar. J. & Youniss, J. (1985). Adolescent self-concept development. In: R. L. Leahy (ed.) The decelopmenr of the self(pp.
247-256). New York: Academic Press.
Spiel, Ch. (1992). Adolescents and parents perceioe each other: Do march and mismatch indicate the quality of relarionshipxj
Paper presented at Vth European Conference on Developmental Psychology, Seville, Spain.
Van der Werff, J. J. (1990). The problem of self-conceiving. In: Bosma, H. & Jackson, S. (Eds) Coping and self-concept in
adolescence (pp. 13-34). Berlin: Springer-Verlag.
Volpe. J. (1981). The development of concepts of self: An interpersonal perspective. In Meacham, J. & Santilli, N. R. (Eds)
Social derelopmenl in youth: Structure and content (pp. 131-144). Basel: Karger.
Youniss, J. & Smollar. J. (1990). Self through relationship development. In Bosma, H. & Jackson, S. (Eds) Coping and selfconcept
in adolescence (pp. 1299148). Berlin: Springer-Verlag.