Self-concept in adolescence. A study of age
and gender differences in groups of normal
and antisocial adolescents
Helene Östgård-Ybrandt & Bengt-Åke Armelius
UMEÅ PSYCHOLOGY REPORTS
No. 3 2003
Umeå Psychology Reports
Acting Editor
Anders Böök
Associate Editors
Bo Molander
Eva Sundin
Ann-Louise Söderlund
Editorial Board
Kerstin Armelius
Anders Böök
Bo Molander
Timo Mäntylä
Eva Sundin
This issue of Umeå Psychology Reports, and recent issues of other departmental reports are available as pdf-files. See
the home page of Department of Psychology (http://www.psy.umu.se/forskning/rapporter/).
Department of Psychology
Umeå University
SE-901 87 Umeå, Sweden
ISSN 1650-8653
Umeå Psychology Reports
No. 3, 2003
1
Self-concept in adolescence. A study of age and gender
differences in groups of normal and antisocial adolescents
Helene Östgård-Ybrandt & Bengt-Åke Armelius
This study examines the self-concept in a group of 277 normal
adolescents with respect to differences in age (12-18 years) and gender
and the interaction between age and gender. As a comparison, 27
antisocial adolescents of similar age participated in the study. The
adolescents completed a questionnaire based on the SASB-model. The
results showed that the normal adolescent self-concept is positive and
that there are no age or gender differences. A more autonomous and
more negative self-concept is found in the antisocial group of
adolescents. The antisocial girls seem to show more self-hate than all
other groups. The conclusion from the study is that there is no evidence
of a chaotic self-concept during normal adolescence, but that an
antisocial way of life might be reflected in a more negative and
autonomous self-concept during adolescence.
Key words: Self-concept, normal, antisocial, adolescence, gender, SASB
The self-concept, and its development over time, has been given a lot of attention
both in psychological theories and in empirical research. A lot of efforts have been
made to define the two most commonly used terms, the self-concept and self-esteem
and difficulties have plagued self-concept research with respect to how these concepts
differ from each other (Skaalvik, 1997; Skaalvik & Bong, 2003). The present study is
concerned with the self-concept issue as operationalized by means of the SASB
system (Structural Analysis of social Behavior) as developed by Benjamin (1974,
1987, 1996a, 1996b). This self-concept is conceptualised in behavioral terms and
expressed as a cognitive perception of the individual’s treatment of him- or herself
(e.g. I like myself very much and welcome and enjoy opportunities to be with myself,
I accuse and blame myself, make myself feel bad, guilty, ashamed, unworthy). The
self-concept in the SASB model is seen as a product of interpersonal interactions
(Sullivan, 1953) and Florsheim, Henry and Benjamin (1996), have added theoretical
assumptions in terms of Bowlby’s attachment theory (Bowlby, 1978). The two
fundamental processes, attachment seeking (affiliation) and exploration (autonomous
differentiation) involved in a child’s relations to nurturing caregivers are in the model
expressed in the two dimensions affiliation and control. These two interpersonal
processes are the basis for the psychological development of the child’s self-concept
and constitute the prototype for the child’s interpersonal relations throughout life.
The normal self-concept includes self-affirmation, active self-love and self-protection
and these normative behaviors must be present most of the time” together with
flexibility “to shift to any point in interpersonal space if the context requires such a
change” (Benjamin, 1993).
The focus of the present paper is on the development of the self-concept in the
adolescence period, the age-span from 12 through 18 years of age, and on possible
gender differences. This is a period in which a lot of changes of the adolescent’s
2
social and interpersonal life occur. In addition, the intellectual capacity increases and
the adolescent becomes more self-conscious.
The adolescence has theoretically been accounted for both as a period of
heightened “storm and stress” with a change in the self-concept in a discontinuous
manner (Arnett, 1999; Blos, 1962; Erikson, 1968; Freud, 1983; Sullivan, 1953) and
as a stage characterized by gradual development of the self-concept rather than by
disruption (Stern, 1985). Results from empirical studies of the self-concept
development through the years of the adolescence period show no consistent pattern
of age differences, some studies suggest stability (Monge, 1973; Savin-Williams &
Demo, 1984), while other studies have found instability (Ellis & Davis, 1982; Scott
& Hoffman, 2002). The inconsistency of empirical results may be dependent on
differences in the variety of instruments used (Skaalvik, 1986).
Studies of gender differences in the development of the self-concept during
adolescence yield quite similar inconclusive results. The results vary from general
gender differences (Bolognini, Plancherel, Bettschart & Halfon, 1996; Harper &
Marshall, 1991; Marsh, 1989), gender differences in specific ages (Monge, 1973),
and to no gender differences found in the self-concept during adolescence (Mboya,
1999). When gender differences are reported a common result is a lower level of selfappreciation for girls than boys. Gender related differences are also found on specific
aspects of the self-concept generally consistent with sex stereotypes, e.g. higher
sociability among girls than boys and a higher appreciation for own achievement
among boys (Dusek & Flaherty, 1981; Monge, 1973).
Factors other than age and gender that are important in the development of the
adolescent self-concept have been associated with a negative self-concept, e.g.
antisocial and criminal behavior (Hay, 2000; Marsh, Parada, Yeung & Healey,
2001), psychological distress with symptoms of depression and suicidal ideation
(Adamson & Lyxell, 1996; Orbach, Mikulincer, Cohen & Stein, 1998; Overholser,
Adams, Lehnert & Brinkman, 1995) and poor parental relations (Dekovic & Meeus,
1997).
Studies on the self-concept in the normal adolescence based on the SASB model
are few and limited to studies of late adolescence. The results of a Swedish study
show that the majority of the youth in a group of adolescents of 18 to 20 years of age
expressed a positive and stable self-concept. The findings indicated a pattern during
adolescence where the control but not the affiliation aspect was the focus of
development (Adamson & Lyxell, 1996).
The SASB model does not include any theoretical assumptions about age or
gender differences in the development of the self-concept. The question of age and
gender differences during adolescence is an important area of research and it seems
important to add these developmental aspects to the SASB model. The purpose of
the present paper is therefore to study the self-concept of adolescents with the main
focus on the self-concept in a group of normal adolescent boys and girls in the agespan 12 through 18 years of age. A small group of antisocial adolescents are included
in the study to examine possible deviations from the pattern found in the normal
adolescent group.
3
Method
Participants
Normal adolescent group. A total of 277 adolescents (131 boys and 146 girls) ranging
in age from 12 to 18 years participated in the study. The mean age was 14.9 years.
The subjects were systematically selected to get an equal number of boys and girls in
the different age groups. However, the final selection had a higher percentage of
adolescents in the early adolescence (12 to 15 years) than in late adolescence. The
adolescents were students in four middle and junior high schools and one high
school in a small town (Umeå) in the north of Sweden. To obtain a representative
distribution with respect to socioeconomic factors, subjects were drawn from schools
in different socio-economics areas and from vocational and academic classes in the
high schools. The students also completed the Youth Self-Report problem scales
(Achenbach, 1991) and were found to have a total score of 40.1 (SD = 19.5), with
scores on anxious/depressed being 5.6 (SD = 4.8), on social problems 2.2 (SD = 2.3),
and on delinquent behavior 3.8 (SD = 2.8), all within the normal range and
equivalent to results from other studies of normal Swedish adolescents (Broberg et
al., 2001).
Antisocial adolescent group. The group of antisocial adolescents consisted of 27
adolescents (8 boys and 19 girls) from the project Psychological problems among
adolescents with antisocial behavior (Armelius & Hägglöf, 1998). They were
between 13 and 19 years of age and the mean age was 15.1 years. Sixty-seven percent
of the adolescents in the group were found in the age range between 12 to 15 years
and this was true for 100% of the boys compared to 53% of the girls. The antisocial
adolescents were in residential care in four different special youth homes. Most of
them were in custody without their parent’s consent, being held under the Swedish
Care of Young Persons Act (LVU). When the care was voluntary, it was based on the
Social Service Act (SOL). The time in custody ranged from 1 to 172 months (M =
38 months). The incidence of psychiatric problems in the group of antisocial
adolescents was considerably higher than the 10 % in the normal Swedish
population between the ages of 12 and 17 years where about 15% of the adolescents
up to 18 years of age are seeking help from a child psychiatric clinic (Statens
offentliga utredningar, 1998). Seventy-nine percent of the antisocial adolescents had
been treated in a child psychiatric clinic and the most common diagnoses for the
group was conduct disorder (88%), major depression (58%), phobia (58%) and
anxiety (25%) defined as in DSM III-R (American Psychiatric Association, 1989).
The incidence of alcohol and drug addiction was 33% and 29%, respectively. The
antisocial group in the study is probably not representative of the group of antisocial
adolescents in general (e.g. extremely high percentage of DSM III-R diagnoses in the
group).
The age distribution for the normal and antisocial adolescent groups is shown in
Table 1. The research ethics committee of the medical faculty at the University of
Umeå has approved the project as a whole.
4
Table 1.
Age distribution in the normal and antisocial adolescent groups
Normal group (n)
Antisocial group (n)
Boys
Girls
Boys
Girls
12
24
27
13
19
18
3
2
14
18
19
3
2
15
17
16
2
6
16
19
27
17
19
27
3
18
15
12
2
Age
19
4
M
14.8
Total M
14.9
14.9
13.9
16.2
15.1
Note. N = 277 (normal group (131 boys and 146 girls)) and N = 27 (antisocial
group (8 boys and 19 girls)). M and total M are expressed in years.
Procedure
We asked teachers in selected classes in four middle and junior high schools and one
high school to distribute consent forms and written descriptions of the study to their
students in order to recruit participants to the normal adolescent group. The
adolescents were given information about the study by research assistants.
Participation was voluntary and contingent upon both parent and student consent.
When the students returned the signed permission form, they were issued with a
battery of questionnaires about self-concept, parental relationships, and mental
health. In some classes the students filled out the questionnaires during normal
school hours, but usually they were filled out at home. The questionnaires were then
returned to the teacher in a sealed envelope and the adolescent was given a cinema
ticket as a gratuity. The participating rate has been estimated by comparing the
actual participation to the number of possible participants in the different classes at
the time of the study. The rate of absence is not known and the attrition is therefore
probably overestimated since it includes adolescents who did not get information
about the study as well as those who chose not to participate or did not get parental
5
consent. The participating rate was for the total group 67% (boys 60% and girls
74%). A higher rate was found for the older adolescents with 82% (boys 81% and
girls 83%) than for the younger adolescents with 57% (boys 50% and girls 67%).
The antisocial adolescents received information about the study from both the
staff at the detention centre and from the two researchers of the project. Their
participation was voluntary and contingent upon both parental and adolescent
consent. Each adolescent was given a closed envelope with questionnaires and
returned them in the same envelope after completing the task. Some of the
adolescents used a computer-based variant of the questionnaires with assistance from
the project staff. A cinema ticket was given as a gratuity. The participation rate for
the total antisocial adolescent group was 66% (boys 60% and girls 68%). About half
of the attrition consisted of those who chose not to take part and the other part was
in severe psychiatric conditions (like psychosis) or unable to understand the
questions (adolescents with mental retardation).
Instruments
Structural Analysis of Social Behavior (SASB). Participants completed the Swedish
long form version of the SASB-Intrex Introject Questionnaire (36 statements) based
on the SASB model (Armelius, Lindelöf & Mårtensson, 1983).
The SASB model was introduced by Benjamin (1974) and will briefly be
summarized here. The model consists of a circumplex that incorporates different
degrees of affiliation and interdependence as describe in Figure 1.
(1) AUTONOMOUS SELF
(8) NEGLECTING SELF
(2) ACCEPTING SELF
(3) LOVING SELF
(7) REJECTING SELF
(6) CRITICIZING SELF
(4) TEACHING SELF
____ Affiliation axis
------ Interdependence axis
(5) CONTROLLING SELF
Figure 1. The Structural Analysis of Social Behavior (SASB) Introject circumplex
(simplified cluster model). The horizontal axis represents the affiliation dimension
(self-love – self-reject) and the vertical axis represents the interdependence dimension
(self-autonomy – self-control).
6
By means of the Intrex questionnaires it is possible to measure the position of the
self-concept in terms of the two dimensions of the model. In the questionnaires the
participants are instructed to rate how well each statement describes their perception
of their own behavior against themselves. The questionnaire items are rated on a
scale from “do not agree” (scale point 0) to “perfect agreement” (scale point 100). It
should be noted that the aim of the SASB is to capture the person’s perception of
him- or herself and not the actual behavior.
All aspects of the SASB-model that are of interest in this study can be organized at
different levels of aggregation. At the lowest aggregated level, there are the 8
individual clusters, which on the next level may be aggregated to the attachment
group of clusters, AG (clusters 2, 3, 4) and the disruptive attachment group of
clusters, DAG (clusters 6, 7, 8). The cluster groups emphasize the affiliation
dimension of the model (Benjamin 1996a).
At the next level of aggregation there are a number of coefficients, which are based
on patterns of clusters. These include the attack (ATK) and the control (CON)
pattern coefficients. Values are expressed in the range from –1.00 to 1.00 and values
< –0.71 and > 0.71 are described as high. The attack coefficients indicate
endorsement of items relative to the horizontal axis expressing affiliation. High
ratings around clusters 6, 7, and 8 (see Figure 1) indicate a high positive attack
pattern or negative affiliation (self-reject). High ratings around clusters 2, 3, and 4
indicate a high negative attack pattern or affiliation (self-love). The control
coefficients indicate endorsement of items relative to the vertical axis expressing
control-autonomy. High ratings around clusters 4, 5, and 6 produce a positive
control coefficient indicating high interdependence and high ratings around clusters
8, 1, and 2 produce a negative control coefficient indicating self autonomy. All levels
of aggregation can in principle be used for all aspects of the SASB model, which gives
an abundance of measures. In the present study we have concentrated our analyses to
the cluster groups and the coefficients.
The test–retest reliability is r = .87 for both the American version (Benjamin,
1987) and the Swedish version (Armelius, 2001) of the SASB Introject
Questionnaires for adults. Factor analyses show that the Swedish translation is
consistent with the model (Armelius & Öhman, 1990; Armelius, 2001). In the
present study satisfactory internal consistency was obtained for the two attachment
clusters. The split-half (correction with Spearman-Brown) for the self-concept was r
= .75 for AG and r = .83 for DAG.
Statistical analyses
The results were analyzed by means of two-way MANOVAs on the different
aggregation levels separately and with the cluster groups combined with the clusters 1
and 5, as suggested by Pincus, Gurtman, and Ruiz (1998). The separate analyses
were undertaken since there are relatively strong dependencies between many of the
dependent variables. When the MANOVA showed at least one significant difference,
univariate ANOVAs were performed on each dependent variable. Bonferroni
corrections were generally applied to adjust for the many comparisons. Since the
antisocial group included only 8 boys we decided to exclude group comparisons
involving antisocial boys on the individual clusters, except for cluster 1 and 5, which
represent the control dimension on the cluster group level of aggregation. We have
shown η2 values to give an idea about the amount of variance accounted for in
addition to the F-values.
7
Results
Self-concept of normal adolescents
In order to answer the question if there are age and/or gender related differences in
self-concept among normal adolescents MANOVAs for clusters, cluster groups and
coefficients with 7 age groups (12-18 years of age) and gender (boys and girls) were
applied to the self-concept ratings. With an alpha level of .05, there were no
significant effects, neither for age (F (48, 1566) = .79, p < .84, η² = .02 for clusters,
F (12, 526) = .67, p < .79, η² = .02 for cluster groups, and F (12, 526) = .63, p < .82,
η² = .01 for coefficients), nor for gender (F (8, 256) = 1.13, p < .34, η² = .03 for
clusters, F (2, 262) = .18, p < .84, η² = .001 for cluster groups, and F (2, 262) = .22,
p < .80, η² = .002 for coefficients,) or the interaction between age and gender (F
(48, 1566) = .84, p < .78, η² = .03 for clusters, F (12, 526) = 1.10, p < .35, η² = .03
for cluster groups, and F (12, 526) = 1.04, p < .42, η² = .02 for coefficients). The
mean for ATK was -.74 (SD = .33) and 78 % of the normal adolescents had a
positive self-concept (ATK < -.71) while the mean of CON was -.02 (SD = .27)
indicating a balance between control and autonomy. A paired t-test (t (276) = 28.1,
p < .001) revealed a significantly higher mean value for AG (M = 67, SD = 15.93)
than for DAG (M = 20, SD = 15.54). The average results are presented in table 2.
Table 2
Mean and standard deviation for the SASB pattern coefficients and clusters in the normal
adolescent group and for boys and girls separately.
Gender
Boys
ATK CON AG DAG Cl 1
M -.75 -.03
SD
Girls
.27
.37
20
55
73
69
Cl 4
Cl 5
Cl 6
Cl 7
Cl 8
62
59
19
15
25
.25 15.2 15.1 16.3 19.8 17.8 16.4 18.9 18.3 16.1 17.0
M -.72 -.00
SD
68
Cl 2 Cl 3
67
21
53
71
69
61
56
22
17
24
.28 16.6 16.0 14.2 21.4 18.7 16.4 18.7 20.2 18.0 14.9
Note. ATK = attack pattern coefficient, CON = control pattern coefficient, AG =
attachment group of clusters, DAG = disrupted attachment group of clusters, Cl =
cluster. N = 277 (131 boys and 146 girls).
In figure 2 the average cluster values of the age groups are plotted in order to
permit a closer look at the self-concept pattern. It is evident that the normal
adolescent group expressed self-acceptance, love and care with a low degree of
hostility and blame in their perceptions of themselves. In fact, only 1 % of the
normal adolescents showed a negative self-concept with a value of ATK > .71. The
cluster curve of these four adolescents was the opposite of the group with a positive
8
self-concept with low values on the positive clusters and high on the negative
clusters.
100
12 yr
90
13 yr
80
14 yr
70
15 yr
60
16 yr
17 yr
50
18 yr
40
30
20
10
0
Autonomous
self
Accepting
self
Loving self
T eaching self
Controlling
self
Criticizing
self
Rejecting self
Neglecting
self
Figure 2. Group means on SASB clusters for self-concept ratings for seven different
age groups of normal adolescents.
The responses to the individual items of the Intrex questionnaire were analyzed in
order to see if there might be differences between age groups and/or gender groups
(ANOVA). The results showed that the younger adolescents of 12-13 years of age
rated themselves higher than older adolescents of 18 years of age on an item
indicating more autonomy “I am happy-go-lucky, content with “here today, gone
tomorrow” ( F (6, 259) = 3.049, p < .01). Girls rated themselves higher than boys on
an item indicating negative affiliation “I am very unsure of myself because I tell myself I
do things all wrong. I feel others can do better” (F (1, 259) = 7.810, p < .01) and boys
rated themselves higher on an item indicating autonomy “I just let important choices,
thoughts, issues, options slip by me without paying much attention” (F (1, 260) = 4.615,
p < .03). There was also a significant interaction effect between age and gender on
the item “I let myself feel glad about and pleased with myself just as I am” (F (6, 258) =
2.417, p ≤.03) indicating that girls of 14-16 years of age rated themselves lower in
positive affiliation than boys during this age period but not during other periods.
Self-concept of the antisocial adolescents
A series of analyses were performed in order to study the difference between the selfconcept of the normal adolescents and the antisocial adolescents. Unfortunately, the
antisocial group was very small and did only allow for certain comparisons. There
were no more than a few age groups present and the distribution of boys and girls in
the different age groups was skewed. It was however possible to compare the normal
and antisocial groups in general and with respect to gender differences.
9
MANOVAs were conducted across the two coefficients ATK and CON, cluster
groups AG and DAG and clusters 1 and 5 to test for differences between the normal
and antisocial groups and for gender differences. The results showed a significant
main effect of groups (F (2, 299) = 16.39, p < .0001, η² = .10 for coefficients, and F
(4, 297) = 4.46, p < .002, η² = .06 for cluster groups), of gender (F (2, 299) = 6.55, p
< .002, η² = .04 for coefficients, and F (4, 297) = 3.41, p < .01, η² = .04 for cluster
groups) and for the interaction effect of groups and gender (F (2, 299) = 5.03, p <
.01, η² = .03 for coefficients, and F (4, 297) = 2.60, p = .04, η² < .03 for cluster).
The univariate ANOVAs on each coefficient, cluster group and individual cluster for
group, gender and the groups and gender interaction are summarised in Table 3.
Table 3
Summary of ANOVAs of SASB clusters, pattern coefficients and cluster groups for the selfconcept ratings in the normal and antisocial adolescent groups
Group
Gender
Group x gender
Variable
F (1, 300)
η²
F (1, 300)
η²
F (1, 300)
η²
Cl 1
3.31
.01
3.04
.01
1.33
.00
Cl 2
.96
.00
3.69
.01
2.28
.01
Cl 3
7.21**
.02
1.09
.00
.88
.00
Cl 4
.64
.00
3.30
.01
2.48
.01
Cl 5
3.49
.01
.73
.00
.03
.00
Cl 6
1.64
.01
16.45***
.05
12.36***
.04
Cl 7
10.84***
.04
4.72*
.02
3.53
.01
Cl 8
7.19**
.02
6.67**
.02
7.92**
.03
ATK
20.12***
.06
12.04***
.04
9.54***
.03
CON
6.14**
.02
3.23
.01
2.02
.01
AG
2.89
.01
3.37
.01
2.36
.01
DAG
7.09**
.02
.04
9.46***
.03
11.03***
Note: * p ≤ .05. ** p ≤ .01. *** p ≤ .001.
10
The main effects of groups were significant for the ATK and CON coefficients and
the cluster group DAG indicating that in general the self-concept of the antisocial
adolescents were more negative and autonomous than that of the normal adolescents.
Only 41 % of the antisocial versus 78 % of the normal adolescents showed a positive
self-concept and a larger part (11 %) of the antisocial adolescent group compared to
the normal group (1 %) showed a negative self-concept. The mean values for the
coefficients, cluster groups and individual clusters for the antisocial adolescent group
are shown in Table 4.
Table 4
Mean scores and standard deviations for the SASB pattern coefficients and clusters in the
antisocial adolescent groups and for boys and girls separately.
Gender
Boys
ATK CON AG DAG Cl 1 Cl 2 Cl 3 Cl 4 Cl 5
M -.64 -.26
SD
Girls
M
SD
Total
M
SD
.24
18
66
75
62
64
52
.24 25.1 10.9 23.4 30.8 25.3 26.9 28.6
-.13 -.07
.69
67
55
42
56
59
54
52
48
Cl 6 Cl 7 Cl 8
9
21
24
6.9 17.5 23.0
43
37
45
.38 23.9 30.0 20.4 31.4 23.9 21.4 26.8 34.7 30.1 30.5
-.28 -.13
58
35
58
64
56
56
49
33
32
39
.64 . 35 24.5 27.8 21.4 31.6 24.1 23.4 26.8 33.0 27.7 29.6
Note. ATK = attack pattern coefficient, CON = control pattern coefficient, AG =
attachment group of clusters, DAG = disrupted attachment group of clusters, Cl =
cluster. N = 27 (8 boys and 19 girls).
The differences in self-concept pattern between the normal and antisocial
adolescents are shown in figure 3, where the average cluster scores are plotted for
the two groups. In figure 3 it can be seen that there is a significantly higher
incidence of self-reject and self-neglect together with less self-love in the group
of antisocial adolescents.
11
80
70
N orm al adolescent group
Antisocial adolescent group
60
*
50
*
40
*
30
20
10
0
Autonom ous Accepting
self
self
Loving
self
Teaching
self
C ontrolling
self
C riticizing
self
R ejecting
self
N eglecting
self
Figure 3. Group means on SASB clusters for self-concept ratings for normal and
antisocial adolescents.
The main effect of gender showed a more negative self-concept (ATK, DAG) for
girls than boys. Since there were no such differences between boys and girls in the
normal group, however, these differences were attributed to differences in the
antisocial adolescent group. The analysis of interaction effects confirmed a more
negative self-concept for girls in the antisocial group than for both boys and girls in
the normal adolescent group and for boys in the antisocial group (ATK, DAG).
The differences in self-concept pattern between the normal and antisocial
adolescent boys and girls are shown in figure 4. In Figure 4 it can be seen that the
antisocial girls tend to be more self-critical (cluster 6) and more self-neglecting
(cluster 8) than all other groups.
12
80
70
Normal boys
Normal girls
Antisocial boys
Antisocial girls
60
50
40
30
20
10
0
Autonomous Accepting
self
self
Loving
self
Teaching
self
Controlling Criticizing
self
self
Rejecting
self
Neglecting
self
Figure 4. Group means for SASB self-concept ratings for normal and antisocial
adolescent boys and girls.
Discussion
In contrast to expectations in many theories proposing a discontinuity in the selfconcept during adolescence we found no differences between normal adolescents in
age groups ranging from 12 through 18 years of age on any of the measures of the
SASB. Thus, this finding adds support to other research (Chubb, Fertman & Ross,
1997; Harter, 1998) and research reviews (Steinberg, 1999; Steinberg & Morris,
2000) which found the self-concept stable over time in the adolescent period.
The absence of differences between age groups on the control dimension is a little
surprising since the ability for internal control could be expected to vary with age.
Normally, the level of external or parental control is higher for young adolescents
and is gradually replaced with internal control until adulthood. Thus, a parallel
development could be expected for the self-concept, but we found no evidence for
such a difference with respect to internalized control. One possible explanation is
that the SASB control dimension does not capture all important aspects of control
behavior in the adolescent period.
Furthermore, the results show no evidence of differences between boys and girls
in the self-concept, only on a few individual items. In contrast to results proposing
more autonomy in the self-concept of the normal adolescent boys (Douvan &
Adelson, 1966) and less self-love in the normal adolescent girls (Harper & Marshall,
1991) our results show no such differences. Similar results, no gender differences,
have been found in groups of late adolescents (high schools and college students) and
in normal and clinical groups of adults examined with the SASB (Armelius, 2001).
The predominantly positive self-concept for both boys and girls is in line with the
general theoretical assumptions of the SASB model (Florsheim, Henry & Benjamin,
1996) assuming that the early parenting are the source for the self-concept
13
development (Bowlby, 1978; Sullivan, 1953). According to these theories the early
“good-enough” mother and father behavior constitute the base for a positive selfconcept and become internalized to similar behavior towards the self regardless of the
gender of the child.
The inconsistency between the results of this study and studies reporting gender
and age differences might be due to the fact that two different aspects of the selfconcept are measured. In SASB the focus is on how the individual relates to himself
and treats himself in behavioral terms, whereas in the other studies the focus is on the
subjective evaluation of the self, the self-esteem. This aspect of the self-concept is
probably more dependent on the present context than the aspect of the self-concept
that is measured with SASB and could for that reason vary with age and gender.
Some support for this interpretation is found in the fact that the girls agreed
significantly more to an item with an obvious evaluative component: “I am very
unsure of myself because I tell myself I do things all wrong. I feel others can do better”. A
further possible explanation for the lack of differences between age and gender
groups in the present study might be that the self-concept of the SASB-model is
insensitive to such differences in adolescence. However, the results clearly show that
there were some differences between the normal and the antisocial adolescents with
respect to the self-concept.
The antisocial adolescents had a more negative and autonomous self-concept and
antisocial girls had much more self-hate than all other groups of adolescents. This
might reflect the pattern of treatment they have received from important others
which has become internalized patterns of attack and neglect in line with the results
suggested in studies by Dekovic and Meeus (1997), Florsheim, Tolan and Gorman –
Smith (1996) and Ruchkin, Eisemann and Hägglöf (1998) and with the theoretical
assumptions connected to the SASB model (Benjamin, 1996a; Henry, 1994). It is
also probable that the negative self-perceptions are reinforced in the adolescents´
actual present relations with their parents. The behavior of the antisocial adolescents
is likely to elicit negative actions from the parents resulting in an increase of the
already negative self- perception of the adolescents.
A contextual explanation to the increased autonomy in the group of antisocial
adolescents may be found in the fact that they were incarcerated in a restricted
environment at the time of the study, which might have increased their need for
autonomy. A higher level of autonomy has been found in groups of adolescents with
problematic behavior (Sussman, McCuller & Dent, 2003) and has been associated
with deficiencies in attachment (Lee & Bell, 2003) and personality (Unnever &
Cornell, 2003).
In general, many of the girls with an antisocial way of life have been exposed to
sexual assaults and physical/psychical abuse (Statens institutionsstyrelse, 2000),
which might well result in a negative self-concept with more self-hate and rejection
of the self. They also suffer from psychological distress due to their present antisocial
way of living and several studies have shown that psychological suffering is closely
related to a negative self-concept (Harper & Marshall, 1991; Bolognini, Plancherel,
Bettschart & Halfon, 1996).
We do not know if the restricted environment, the sexual and physical abuses and
psychological distress have similar effects on the self-concept of antisocial boys and
girls. Interesting enough, the boys in the antisocial group showed self-love to nearly
the same extent as the boys and girls in the normal adolescent group. The selfconcept of the antisocial boys may be elevated by compensatory self-enhancement
and self-protection processes aiming at avoidance of loss of self-esteem (Baumeister,
14
Boden, & Smart, 1996). Perceptions of superiority in the group of the antisocial
boys may have been expressed as perceived self-love. Bushman and Baumeister
(2002) found that a sense of superiority is connected to aggressive antisocial
behavior. The self-concept of the antisocial boys may be gratified and reinforced by
the peer-group to a larger extent than the self-concept of the antisocial girls (Hay,
2000).
The conclusion drawn from the results from the normal adolescent group, that
there are very small differences in the self-concept between boys and girls in the age
range 12-18 years of age, seems to be reliable since it is based on a relatively large and
representative sample. However, the results from the comparison between the normal
and antisocial adolescent groups do not allow for general conclusions but should be
considered as an interesting result that needs further empirical study.
15
References
Achenbach, T. M. (1991). Manual for the youth self-report and 1991 profile.
Burlington: University of Vermont, Department of Psychiatry.
Adamson, L., & Lyxell, B. (1996). Self-concept and questions of life: identity
development during late adolescence. Journal of Adolescence, 19, 569-582.
Arnett, J. J. (1999). Adolescent storm and stress, reconsidered. American Psychologist,
45, 317-326.
American Psychiatric Association. (1989). Diagnostic and statistical manual of mental
disorders (3rd. ed.). Stockholm: Pilgrim press.
Armelius, B-Å., & Hägglöf, B. (1998). Psykiska störningar hos ungdomar med sociala
problem [Psychological problems among adolescents with antisocial behavior].
Sweden: University of Umeå, Department of Psychology.
Armelius, K., Lindelöf, I-S. & Mårtensson, B. (1983). Structural Analysis of Social
Behavior. Tips No.5. Sweden: University of Umeå, Department of Applied
Psychology.
Armelius, K., & Öhman, K. (1990). Factor analysis of the Swedish introject version of
SASB. Tips No.36. Sweden: University of Umeå, Department of Applied
Psychology.
Armelius, K., (2001). Reliabilitet och validitet för den svenska versionen av SASBsjälvbildstestet [Reliability and validity for the Swedish version of SASB-introject
test]. Unpublished manuscript, Sweden: University of Umeå, Department of
Psychology.
Baumeister, R. F., Boden, J. M., & Smart, L. (1996). Relation of threatened egotism
to violence and aggression: the dark side of high self-esteem. Psychological Review,
103, 5-33.
Benjamin, L. S. (1974). Structural Analysis of Social Behavior. Psychological Review,
81, 392- 425.
Benjamin, L. S. (1987). SASB Short form users manual. Utah: University of Utah
Department of Psychology.
Benjamin, L. S. (1993). Interpersonal diagnosis and treatment of personality disorders.
New York: Guildford Press.
Benjamin, L. S. (1996a). A clinician-friendly version of the interpersonal circumplex:
Structural Analysis of Social Behavior (SASB). Journal of Personality Assessment,
66, 248-266.
Benjamin, L. S. (1996b). Introduction to the special section on Structural Analysis of
Social Behavior. Journal of Consulting and Clinical Psychology, 64, 1203-1212.
Blos, P. (1962). On adolescence. New York: The Free Press.
Bolognini, M., Plancherel, B., Bettschart, W., & Halfon, O. (1996). Self-esteem and
mental health in early adolescence: development and gender differences. Journal of
Adolescence, 19, 233-245.
Broberg, A. G., Ekeroth, K., Gustafsson, P. A., Hansson, K., Hägglöf, B., Ivarsson,
T., & Larsson, B. (2001). Self-reported competencies and problems among
Swedish adolescents: A normative study of the YSR. European Child and
Adolescent Psychiatry, 10, 186–193.
Bowlby, J. (1978). Attachment and loss, Vol. 1: Attachment. New York: Penguin
Books.
Bushman, B. J., & Baumeister, R. F. (2002). Does self-love or self-hate lead to
violence? Journal of Research in Personality, 36, 543-545.
16
Chubb, N. H., Fertman, C-I., & Ross, J. L. (1997) Adolescent self-esteem and locus
of control: a longitudinal study of gender and age differences. Adolescence, 32,
113-129.
Dekovic, M., & Meeus, W. (1997). Peer relations in adolescence: effects of
parenting and adolescents self-concept. Journal of Adolescence, 20,163-176.
Douvan, E., & Adelson, J. (1966). The adolescent experience. New York: John
Wiley and Sons.
Dusek, J. B., & Flaherty, J. F. (1981). The development of self-concept during
adolescent years. Monographs of the Society for Research in Child
development, 46(4, serial No. 191).
Ellis, D. W., & Davis, L. T. (1982). The development of self-concept boundaries
across the adolescent years. Adolescence, 17, 695-710.
Erikson, E. H. (1968). Identity: Youth and crisis. New York: Norton.
Florsheim, P., Tolan, P., & Gorman-Smith, D. (1996). Family processes and risk for
externalizing behavior problems among African American and Hispanic boys.
Journal of Consulting and Clinical Psychology, 64, 1222-1230.
Florsheim, P., Henry, W. P., & Benjamin, L. S. (1996). Integrating individual and
interpersonal approaches to diagnosis: The Structural Analysis of Social Behavior
and Attachment theory. In Kaslow, F. W. (Ed.), Handbook of relational diagnosis
and dysfunctional family patterns (pp.81-101). New York: John Wiley and Sons.
Freud, A. (1983). Jaget och dess försvarsmekanismer [The ego and the mechanisms of
defence]. Stockholm: Natur och Kultur.
Harper, J. F., & Marshall, E. (1991). Adolescent´s problems and their relationship to
self-esteem. Adolescence, 26, 799-808.
Harter, S. (1998). The development of self-representations. In W. Damon (Series
Editor) & N. Eisenberg (Volume Editor), Handbook of child psychology: Vol. 3:
Social, emotional, and personality development (5th ed). New York: Wiley.
Hay, I. (2000). Gender self-concept profiles of adolescents suspended from high
school. Journal of Child Psychology and Psychiatric, 41, 345-352.
Henry, W. P. (1994). Differentiating normal and abnormal personality: an
interpersonal approach based on the Structural Analysis of Social Behavior. In
Strack, S. & Lorr, M. (Eds), Differentiating normal and abnormal personality. New
York, US: Springer Publishing.
Lee, J-M., & Bell, J. (2003). Individual differences in attachment-autonomy
configurations: linkages with substance use and youth competencies. Journal of
Adolescence, 26, 347-361.
Marsh, H. W. (1989). Age and sex effects in multiple dimensions of self-concept:
preadolescence to early adulthood. Journal of educational Psychology, 81, 417-430.
Marsh, H. W., Parada, R. H., Yeung, A. S., & Healey, J. (2001) Aggressive school
troublemakers and victims: A longitudinal model examining the pivotal role of
self-concept. Journal of educational psychology, 93, 411-419.
Mboya, M. M. (1999). Multiple dimensions of adolescent self-concept. Relation
with age, gender and scholastic measures. School Psychology International, 20, 388398.
Monge, R. H. (1973). Developmental trends in factors of adolescent self-concept.
Developmental Psychology, 8, 382-393.
Orbach, I., Mikulincer, M., Cohen, O., & Stein, D. (1998). Self-representation of
suicidal adolescents. Journal of Abnormal Psychology, 107, 435-439.
17
Overholser, J., Adams, D. M., Lehnert, K. L., & Brinkman, D. C. (1995). Selfesteem deficits and suicidal tendencies among adolescents. Journal of the American
Academy of Child and Adolescent Psychiatry, 34, 919-928.
Pinkus, A. L., Gurtman, M. B., & Ruiz, M. A. (1998). Structural Analysis of Social
Behavior (SASB): Circumplex analyses and structural relations with the
interpersonal circle and the five-factor model of personality. Journal of Personality
and Social Psychology, 74, 1629-1645.
Ruchkin, V. R., Eisemann, M., & Hägglöf, B. (1998). Parental rearing and problem
behaviours in male delinquent adolescents v s controls in Northern Russia. Social
Psychiatry and Psychiatric Epidemiology, 33, 477-482.
Savin-Williams, R. C., & Demo, D. H. (1984). Developmental change and stability
in adolescent self-concept. Developmental Psychology, 20, 1100-1110.
Scott, A. B., & Hoffman, J. P. (2002). The dynamics of self-esteem: A growt-curve
analysis. Journal of Youth and Adolescence, 31, 101-113.
Skaalvik, E. M. (1986). Sex differences in global self-esteem. A research review.
Scandinavian Journal of Educational Research, 30, 167-179.
Skaalvik, E. M. (1997). Issues in research on self-concept. In: Maehr, M., and
Pintrich, P. R. (eds.), Advances in Motivation and Achievement (Vol. 10), 51-97.
JAI Press, New York,
Skaalvik, E. M., & Bong, M. (2003). Academic self-concept and self-efficacy: How
different are they really? Educational Psychology Review, 15, 1-40.
Statens institutionsstyrelse. (2000). Årsrapport ADAD 97 [Yearly report ADAD 97]
(SIS publication 1/00), Sweden: Author. Statens offentliga utredningar. (1998).
Det gäller livet. Stöd och vård till ungdomar med psykiska problem [Life is at stake.
Support and treatment for youth with psychiatric problems] (ISSN: 0375250X;
1998:31). Stockholm: Author.
Steinberg, L. (1999). Adolescence. (5th ed.). New York: McGraw-Hill.
Steinberg, L., & Morris, A. S. (2000). Adolescent development. Annual-Review-ofPsychology, 52, 83-110.
Stern, D. N. (1985). Spädbarnets interpersonella värld ur psykoanalytiskt och
utvecklingspsykologiskt perspektiv [The interpersonal world of the infant: A view
from psychoanalysis and developmental psychology]. Stockholm: Natur och
Kultur.
Sullivan, H. S. (1953). The interpersonal theory of psychiatry. New York: Norton.
Sussman, S., McCuller, W. J., & Dent, C. W. (2003). The associations of social selfcontrol, personality disorders, and demographics with drug use among high-risk
youth. Addictive Behaviors, 28, 1159-1166.
Unnever, J. D., & Cornell, D. G. (2003). Bullying, self-control and ADHD. Journal
of Interpersonal Violence, 18, 129-147.
Umeå University S-901 87 Umeå.
Phone +46 90 786 50 00, Fax +46 90 786 99 95
ISSN 1650-8653