Professional Documents
Culture Documents
Adolescent Parental Attachment and Acade PDF
Adolescent Parental Attachment and Acade PDF
Adolescent Parental Attachment and Acade PDF
In this study, the links between adolescents’ attachment to parents and academic
motivation and performance were examined while considering problem behaviors
and perceived teacher support as potential mediators of those links. Participants
were 121 early adolescents who completed the Inventory of Parent and Peer Attach-
ment (IPPA) to assess the quality of attachment to their mothers and fathers.
Adolescents’ attachment to both parents was positively associated with academic
motivation. These significant links were mediated by adolescents’ perceptions of
support from teachers. Results are discussed in light of the different mechanisms that
can relate attachment quality to adolescents’ academic motivation and performance.
1
The research described in this article was supported by grants from the Social Sciences and
Humanities Research Council of Canada (#410-98-0892).
2
Correspondence concerning this article should be addressed to Stéphane Duchesne or
Simon Larose, Faculty of Education Sciences, Laval University, Laval, Québec, Canada G1K
7P4.
1501
& Eccles, 1998; Seidman, Allen, Aber, Mitchell, & Feinman, 1994). For a
significant number of adolescents, this transition coincides with an increase in
internalized and externalized problems (Achenbach, Howell, Quay, &
Conners, 1991; Kazdin, 1993); a decrease in intrinsic motivation, self-
confidence, and perceived academic competence (Goodenow, 1993; Harter,
1996; Wentzel, 1994, 1998); and a deterioration in academic performance
(Barber & Olsen, 2004; Eccles et al., 1993).
The goal of the present study is to determine whether attachment quality
with both parents is associated with adolescents’ academic motivation and
performance during the first year in high school (i.e., Grade 7) and to test the
role of problem behaviors (i.e., low internalized and externalized problems)
and perceived teacher support as potential mediators of those links. In our
view, attachment theory is a promising perspective from which to understand
the personal mechanisms linked to academic processes and outcomes during
the first year in high school because it highlights the critical importance of
social exploration, management of distress, and development of trust rela-
tionships (Feeney & Noller, 1996), three developmental tasks that might play
a crucial role during school transitions.
drug use, vandalism (Noom et al., 1999), and aggressive behavior (Laible
et al., 2000; Noom et al., 1999); and internalized problems, such as anxiety
and depressive mood (Allen, Moore, Kuperminc, & Bell, 1998; Nada-Raja,
McGee, & Stanton, 1992). On the other hand, the association between these
problem behaviors and academic outcomes has been well documented in
the literature. For instance, some studies have shown that students who
control their impulsiveness in class appear to make a better adjustment to
school, whereas those who manifest higher levels of anxiety and depression
perceive themselves as lacking in academic ability (e.g., Cole, Martin, Peeke,
Seroczynski, & Fier, 1999).
Integrating these two lines of research with that of attachment theory, it
was expected that internalized and externalized problem behaviors would
mediate the link between poor attachment quality and academic motivation
and performance during the high school transition. Children lacking a secure
base, instead of turning to parents or other persons whom they trust, would
seek other means to regulate their negative emotions. The exhibition of
internalized (e.g., sadness) and externalized (e.g., aggressiveness) problem
behaviors could constitute a way for children to express their needs for
security. Moreover, those behaviors would increase the risk that adolescents
entering high school will be rejected by their peers and perceived negatively
by their teachers (Kupersmidt, Coie, & Dodge, 1990), thereby affecting their
academic motivation and performance.
On the contrary, attachment quality to parents would provide the ado-
lescent with a context that is favorable to the healthy management of nega-
tive emotions. This function could allow the adolescent to better tolerate
strange and potentially stressful situations (e.g., starting high school) and to
use behavioral strategies adapted to these situations (e.g., talking to parents
or teachers about their negative emotions). As a result, attachment quality to
parents might provide protection against the development of internalized and
externalized behavioral problems in school, thereby facilitating adolescents’
academic motivation and performance.
Attachment to parents has been related also to adolescents’ perceptions of
their affective relationships with teachers and academic counselors (Cotterell,
1992; Larose & Boivin, 1998; Soucy & Larose, 2000), and perceptions of
affective relationships in the school environment to adolescents’ academic
adjustment (Eccles et al., 1993; Moos, 1991; Roeser & Eccles, 1998; Roeser,
Eccles, & Sameroff, 1998; Ryan, Stiller, & Lynch, 1994; Wentzel, 1998).
Following the same lines of reasoning, it is expected that perceptions of
teacher support will mediate the relation between attachment and academic
motivation and performance.
According to attachment theory, security deriving from the relationship
with parents encourages social exploration (Bowlby, 1969). In a period of
ADOLESCENT PARENTAL ATTACHMENT AND ACADEMIC PERFORMANCE 1505
calm, this security could predispose children to move away from their
parents, while at the same time ensuring that the parent would be available
if needed. In adolescence, this secure exploration could foster social
skills (e.g., ability to develop significant ties to others) and favor the devel-
opment of harmonious relationships with peers and adults other than
parents. This function of attachment, therefore, suggests that attachment
quality to parents can lead adolescents to perceive more support in rela-
tionships with teachers, which in turn should facilitate their academic moti-
vation and performance. Conversely, adolescents who exhibit poor-quality
attachment to parents will be more likely to have a negative perception of
teacher support and to experience adjustment problems in the first year of
high school.
Method
Procedure
Participants
3
For the great majority of adolescents, the educational system in Québec involves changing
schools between Grade 6 and Grade 7. Few adolescents change school before Grade 6 or after
Grade 7. The second important transition generally occurs between the end of high school
(Grade 11) and the first year of college. This situation, which is specific to the Québec system,
accounted for our decision to collect data at the beginning of Grade 7 in September and October
1998.
ADOLESCENT PARENTAL ATTACHMENT AND ACADEMIC PERFORMANCE 1507
Measures
the disruptive behavior and aggressiveness items. The second factor explains
32.4% of the variance (eigenvalue = 2.52), with positive loadings varying from
.77 to .81 on disruptive behavior and aggressiveness, and nonsignificant
loadings (< .30) on the social withdrawal and sadness items. Consequently,
two dimensions were computed as mean scores (externalized behaviors and
internalized behaviors) and were used as two distinct mediator variables.
Coefficient alphas for those mean scores were .86 and .78 for externalized
problem behaviors and internalized problem behaviors, respectively.
Perceived teacher support. Adolescents’ perceptions of support provided
by their mathematics and French teachers were assessed with the personal-
ization subscale of the Individualized Classroom Environment Questionnaire
(ICEQ; Fraser & Fisher, 1986). This subscale, which consists of five items
rated on a 5-point scale, evaluates the perceived availability and accessibility
of a particular teacher (e.g., “The mathematics [French] teacher talks with
each student”; “The mathematics [French] teacher takes a personal interest in
each student”; “The mathematics [French] teacher considers students’ feel-
ings”). Reliability and validity of the ICEQ are well documented (see Fraser,
1987, 1991). In the present study, coefficient alpha for the personalization
score was .81.
Perceived learning environment. To take into account the expected influ-
ence of the perceived learning environment on adjustment, adolescents com-
pleted two scales of the short version of the Learning Environment Inventory
(LEI; Anderson & Walberg, 1972). These scales cover dimensions that pre-
viously have been related to high school adjustment (Walberg & Anderson,
1968, 1972). Those dimensions are democracy (4 items; e.g., “Decisions in the
class are made by everyone”; “Some students impose their will on the whole
class”) and subject matter difficulty (4 items; e.g., “The course is too easy for
many of the students”; “In this class, the material is difficult”).
Items on this questionnaire were rated on a 4-point scale ranging from
1 (disagree completely) to 4 (agree completely) and referred to the French and
mathematics class environments. Substantial evidence of construct and pre-
dictive validity of the scales has been presented in a number of previous
studies (Anderson & Walberg, 1972; Walberg & Anderson, 1972). In the
current study, coefficient alphas were .71 and .66 for democracy and subject
matter difficulty, respectively.
Academic motivation and academic performance. Adolescents’ motivation
was assessed using the Academic Motivation Scale (Keeves, 1974). This
self-report measure consists of 20 items rated on a 5-point scale ranging from
1 (disagree completely) to 5 (agree completely) and provides an estimate of
students’ motivation and interest in school in general (e.g., “I always work
hard at school”; “I want to go as far as possible in school”; “I rarely push
myself at school,” reverse-scored). The scale has excellent metric qualities,
ADOLESCENT PARENTAL ATTACHMENT AND ACADEMIC PERFORMANCE 1509
Results
The results are presented in three steps. First, ANOVAs were computed to
examine gender differences in attachment, behavior problems, perceived
teacher support, and academic motivation and performance. Second, partial
correlations were computed between attachment quality, behavior problems,
perceived teacher support, and academic motivation and performance, con-
trolling for perceptions of the learning environment (i.e., democracy, subject
matter difficulty). Finally, the presence of mediating effects was tested with
two sets of regression analyses in order to verify whether the presumed link
between attachment (predictor) and academic outcomes (criterion) was
explained by the behavior problems of the adolescents (mediator) and their
perceptions of teacher support (mediator).
Gender Differences
Table 1
Boys Girls
M SD M SD
Attachment to mother 3.90 0.80 3.93 0.83
Attachment to father 3.72 0.75 3.40 0.90
Externalized behaviors 3.10 1.72 1.77 0.98
Internalized behaviors 3.14 1.09 3.12 1.45
Perceived teacher support 3.02 0.91 3.14 0.81
Academic motivation 3.44 0.80 3.76 0.58
Academic performance 66.46 11.64 73.67 12.43
Note. Attachment and perceived teacher support were rated on a 5-point scale
ranging from 1 to 5 (5 = positive perceptions). Problem behaviors were rated on a
7-point scale ranging from 1 to 7 (7 = high level of behavioral problems). Academic
motivation was rated on a 5-point scale ranging from 1 to 5 (5 = high level of
motivation). Academic performance was rated on a 100-point scale ranging from 1 to
100.
suggests that girls are less disruptive and perform better at school than do
boys during the elementary and high school years.
1 2 3 4 5 6 7 8
1. Democracy —
2. Subject matter difficulty -.26** —
3. Adolescents’ attachment to mothers .39*** -.01 —
4. Adolescents’ attachment to fathers .22* -.11 .70*** —
5. Externalized behaviours -.23* .08 -.13 -.10 —
6. Internalized behaviours .01 .23* -.16 .01 -.12 —
7. Perceived support from teacher .52*** -.23* .26** .28** -.24** -.05 —
8. Academic motivation .29** -.09 .29** .28** -.57*** -.21** .33*** —
9. Academic performance .15 -.07 .16 .07 -.49*** -.43*** .18* .53***
Note. Correlations among attachment, problem behaviors, perceived teacher support, academic motivation, and academic perfor-
mance were performed while controlling for the variance explained by quality of the learning environment (democracy and subject
matter difficulty).
*p < .05. **p < .01. ***p < .001.
ADOLESCENT PARENTAL ATTACHMENT AND ACADEMIC PERFORMANCE
1511
1512 DUCHESNE AND LAROSE
Mediating Effects
Sobel’s test produces a Z score that can be used to evaluate whether the
mediation path is greater than 0 when the direct independent–dependent path
is taken into account. This score is derived from the ratio of the betas
obtained in the regression equations of two differing paths of the mediational
model (independent → mediator path; mediator → dependent path) to the
standard error of those betas. By taking into account this last statistical
parameter, it is possible to explain why in certain models of mediation,
reduction of the beta of the attachment variable (Equation B) is not signifi-
cant, even if it appears to be more important than those reported in other
equations. The Z score that results from Sobel’s test is significant (one-tailed
significance levels) from 1.65 ( p < .05; see details in Baron & Kenny, 1986).
Sobel’s formula (Baron & Kenny, 1986) is as follows:
a∗b 2 2
ε √b2 sa2 + + sa sb
a 2 sb2
Discussion
The main objective of the present study was to examine whether class-
room behaviors and perceptions of teacher support mediate associations
between attachment quality and adolescents’ academic motivation and per-
formance in the first year of high school. Adolescents’ perceptions of teacher
support had a partial mediating effect of the link between adolescents’ attach-
4
Analyses were conducted separately for mother and father. Patterns of results do not differ
according to the attachment figure, which suggests that, at least in the present study, they can be
generalized to relationships with mother and father.
1514 DUCHESNE AND LAROSE
Table 3
B SE B b t R2 Sobel’s test
Equation A .18***
Attachment .30 .09 .32 3.42***
Academic motivation with mediator
Equation B .23*** Z = 2.42
Attachment .23 .09 .24 2.53**
Perceptions of teacher .24 .09 .29 2.74**
support
Note. Equations were performed while controlling for variance explained by democ-
racy.
**p < .01. ***p < .001.
ment and their academic motivation. This partial mediating relationship was
identified as significant after controlling for perceived democracy in learning
environment.
This finding suggests that adolescents with high-quality attachment to
parents become more willing to explore their school environment with con-
fidence, which may lead them to develop more positive perceptions of
support from their teacher, thus fostering academic motivation. It also sup-
ports the idea that a positive working model of others derived from a secure
attachment relationship may lead to a better appreciation of the academic
support provided by teachers. However, it should be kept in mind that
perceived support from teachers cannot completely explain the relationship
between attachment to parents and academic motivation, thus suggesting
that other mediators could play an important role.
Contrary to our expectations, externalized and internalized behaviors did
not mediate the link between attachment and academic motivation and per-
formance in adolescents. In the case of externalized behaviors, the absence of
mediation might be attributable to common variance with perceptions of
democracy in the classroom, for which we have controlled. Indeed, data
reported in Table 2 indicate that externalized behaviors were negatively asso-
ciated with democracy (r = -.23, p < .05). Therefore, it is possible that the
mediating potential of problem behaviors was attenuated by the fact that
ADOLESCENT PARENTAL ATTACHMENT AND ACADEMIC PERFORMANCE 1515
these behaviors were already associated with this control variable. The
finding that internalized behaviors did not play a mediating role might be
explained by the fact that this category of behavioral problems is more
difficult to detect in a structured context, such as the one prevailing in the
classroom. In these circumstances, this behavior would be assessed better by
mothers and fathers than by teachers (Gagnon, Vitaro, & Tremblay, 1992) or
by using sociometric measures.
No significant association was detected between attachment and aca-
demic performance. This lack of results, combined with the fact that
attachment is related to academic motivation, suggests that attachment to
parents is likely to be related to more subjective aspects of school adjust-
ment. It may be possible that a secure attachment relationship with parents
contributes to the development of adolescents’ positive perceptions of their
capacities to learn, to make friends, and to develop positive relationships
with teachers, which, in turn, would be positively associated with academic
performance.
Adolescents’ attachment was perceived to be of higher quality with
regard to mothers than to fathers. The transition to high school coincides
with important physical and social changes for boys and girls that charac-
terize their transition to puberty (Brooks-Gunn & Reiter, 1990). These
changes might lead adolescents to structure their attachment behaviors
around their mothers, especially since the latter would be perceived to be
the best persons to provide comfort and support during this developmental
time. Thus, adolescents’ search for closeness, combined with their mothers’
sensitivity, likely would contribute to explaining the stronger attachment
perceived by adolescents with regard to mothers. Further studies might
help to confirm this idea.
On the other hand, it was previously pointed out that the nature of the
affective relationships between fathers and adolescents is different from that
characterizing the mother–adolescent relationship. It seems relevant to con-
sider the main contexts in which attachment is most likely to be expressed. It
is possible that adolescents’ attachment with fathers is defined better when it
is linked with a particular context (e.g., participation in recreational activi-
ties; discussions about future directions; social and political discussions).
Further studies would help to support these hypotheses.
Finally, girls showed fewer externalized behavior problems and higher
academic performance than did boys. This finding is consistent with those
that have been reported in previous studies (e.g., Cole et al., 1999; Vitaro &
Gagnon, 1994) and suggests that girls, by being less aggressive and perform-
ing better at school than boys, are better equipped to adjust to a new school
context in which expectations and demands are higher than those they
encountered in the past.
1516 DUCHESNE AND LAROSE
References
Achenbach, T. M., & Edelbrock, C. (1983). Manual for the Child Behavior
Checklist and Revised Child Behavior Profile. Burlington, VT: University
of Vermont.
Achenbach, T. M., Howell, C. T., Quay, H. C., & Conners, C. K. (1991).
National survey of problems and competencies among four- to sixteen-
year-olds. Monographs for the Society of Research in Child Development,
56(3, Serial No. 225).
ADOLESCENT PARENTAL ATTACHMENT AND ACADEMIC PERFORMANCE 1517