School competence in young adolescence: Links to early attachment relationships beyond concurrent self-perceived competence and representations of relationships

Abstract

Predictive associations of infant attachment to mothers and fathers with later school functioning, beyond the contribution of contemporaneous representations of relationships and circumstances of caregiving, were examined in 66 young adolescents who were raised in infancy in Israeli kibbutzim with collective sleeping. The Strange Situation Procedure was used to evaluate early attachment to mother and father, the Separation Anxiety Test was used to assess contemporaneous representation of relationships, and teachers’ reports evaluated school functioning. Circumstances of caregiving included parental reports of quality of marital relations and a change from collective sleeping to home sleeping for children. Results showed that infant attachment to mother, but not to father, contributed significant additional variance to the prediction of children’s scholastic skills and emotional maturity beyond the contribution of concurrent representations of relationships and changes in circumstances of caregiving. The results support the secure base construct as an organising concept of longitudinal investigations of attachment.

Full text

School Competence and Early Attachment
1
Running Head: SCHOOL COMPETENCE AND EARLY ATTACHMENT
School Competence in Young Adolescence: Links to Early Attachment Relationships beyond
Concurrent Self Perceived Competence and Representations of Relationships
Ora Aviezer
Oranim Teachers College, Israel
Abraham Sagi
University of Haifa, Israel
Gary Resnick
WESTAT, USA
Motti Gini
University of Haifa, Israel
International Journal of Behavioral Development 26:397-409, 2002

School Competence and Early Attachment
2
Abstract
Predictive associations of infant attachment to mothers and fathers to later school functioning,
beyond the contribution of contemporaneous representations of relationships and
circumstances of caregiving, were examined in 66 young adolescents who were raised in
infancy in Israeli kibbutzim with collective sleeping. The Strange Situation Procedure was
used to evaluate early attachment to mother and father, Separation Anxiety Test was used to
assess contemporaneous representation of relationships and teachers’ reports evaluated school
functioning. Circumstances of caregiving included parental reports of quality of marital
relations and a change from collective sleeping to home sleeping for children. Results showed
that infant attachment to mother, but not to father contributed significant additional variance
to the prediction of children’s scholastic skills and emotional maturity beyond the
contribution of concurrent representations of relationships and changes in circumstances of
caregiving. The results support the secure base construct as an organizing concept of
longitudinal investigations of attachment.

School Competence and Early Attachment
3
School Competence in Young Adolescence: Links to Early Attachment Relationships beyond
concurrent self perceived competence and representations of relationships
Effective performance at school is among the most salient developmental tasks in
childhood and adolescence. It involves academic and social competencies as well as
behavioral conduct (Masten, Coatsworth, Neemann, Gest, Tellegen & Garmezy, 1995), and it
is influenced by the affective atmosphere between parents and children. Empirical evidence
indicates that appropriate parental involvement and disciplinary measures (Dornbusch, Ritter,
Leiderman, Roberts & Fraleigh, 1987; Steinberg, Lamborn, Darling, Mounts, & Dornbusch,
1994; Wentzel, Feldman & Weinberger, 1991) as well as perceived parental supportiveness
(e.g. Toth & Cicchetti, 1996; Wentzel, 1998) are associated to school oriented interests,
academic achievement and general school adaptation. In addition, adolescents’ school
adjustment is related to the extent that their family milieu is sensitive to their changing needs
(Eccles, Lord & Buchanan, 1996), while self-perceived attachment to parents is related to
self-perceived competence in transition to junior high school (Papini & Roggman, 1992) and
to academic motivation in high school (Learner & Kruger, 1997).
Such connections between affective home environment and school functioning in
childhood and adolescence are congruent with the secure base construct (Bowlby, 1988) that
views parent-child relations as a basis for children’s adjustment at all ages because the
representational models that mediate developmental continuity are grounded in early
attachment experiences. Thus, experiencing early secure attachment relationships allows
children to develop a sense of autonomy that enables them to seek new information freely and
explore away from the safety of the relationship. Attachment insecurity, however, leads
children to ineffective exploration (Bretherton, 1997) because of their ways of directing
attention and behavior, which restrict the type of information that is made available to them

School Competence and Early Attachment
4
through memory or attentiveness to the environment (Main, Kaplan & Cassidy, 1985). Hence,
the secure base construct has been postulated to be a useful organizational anchor for
understanding the balance between attachment and exploration in the study of long-term
adaptive consequences of early attachment patterns (Waters & Cummings, 2000).
Several prospective studies have found predictive associations between infant
attachment relations and a variety of salient developmental adaptations in early and middle
childhood (e.g. Arend, Gove & Sroufe, 1979; Freitag, Belsky, Grossmann, Grossmann,
Scheuerer-Englisch, 1996; Oppenheim, Sagi, & Lamb, 1988; Sroufe, Carlson & Shulman,
1993; Sroufe, Egeland & Kreutzer, 1990) as well as in late adolescence (Roisman, Bahadur &
Oster, 2000). Such associations correspond to a general expectation that individual differences
in attachment security predict later adjustments across developmental domains (Weinfield,
Sroufe, Egeland &Carlson, 1999). The present research set out from a similar prospective
approach. It focused on school adjustment in early adolescence and examined the predictive
contribution of early attachment relations to school functioning.
Among long-term adaptations that are pertinent to school adjustment are cognitive and
metacognitive functioning which were postulated to be different for individuals who
experienced different quality of early attachment security (Main, 1991; Moss, parent,
Gosselin & Dumont, 1993). Most studies of cognitive functioning associated with attachment
quality were carried out with infants and toddlers. Thus, for example, secure infants were
more free to attend to the environment compared to resistant infants who spent more time
monitoring the mother (Cassidy, 1986). Also the exploration of secure toddlers was more
independent and was less restricted (Hazen & Durrett, 1982), whereas resistant children
showed less positive affect and task persistence during play (Frodi, Bridges & Grolnick,
1985). In addition, mothers of secure children were rated higher on supportive presence and

School Competence and Early Attachment
5
quality of assistance in collaborative problem-solving tasks (Frankel & Bates, 1990; Matas,
Arend & Sroufe, 1978).
Moreover, Main (1991) reported pilot data that linked early attachment security to
higher metacognitive functioning in children 6 and 10 years old. Likewise, higher level of
metacognitive competence was displayed by secure preschoolers during a collaborative task
with mother and stranger, which was paralleled with their mothers’ higher cognitive
sensitivity compared to mothers of insecure children (Moss, Gosselin, Parent, Rousseau &
Dumont, 1997). Based on these findings, the lower cognitive and metacognitive functioning
of insecure children was explained in terms of fewer resources for exploratory activity and
metacognitive processes because of their need to monitor the caregiver’s accessibility (Main,
1991; Moss et al., 1997).
Previous assessments of the links between early attachment representations and school
adaptations in adolescence have been rather scarce. Jacobsen and Hofmann found that
adolescents’ cognitive functioning (Jacobsen, Edelstein & Hofmann, 1994) as well as
teachers’ rated school competence and Grade Point Average (Jacobsen & Hofmann, 1997)
were significantly influenced by attachment representations at age seven after IQ, attention
problems and perspective taking skills were controlled for. Though these findings are
consistent with the secure base construct of attachment theory, it is necessary to consider
infancy attachment in order to evaluate the connections of individual differences in early
relationships to individual differences in adolescence.
Another prospective longitudinal study (Teo, Carlson, Mathieu, Egeland & Sroufe,
1996) constructed early psychosocial history as a composite variable that indicates child’s
adaptation in the first three years, including infant-mother attachment. This composite
variable along with cumulative maternal life stress and child’s earlier socioemotional school
adjustment were used to predict math and reading achievement at age 16. Impressive

School Competence and Early Attachment
6
associations between early history and high school achievement were found after controlling
for IQ and earlier school achievement. However, these findings may result from indirect paths
of influence in which early attachment relations are linked with factors such as home
atmosphere that in turn supports school performance (Weinfield et al., 1999). Alternatively,
intercorrelated features of early experience, such as parental sensitivity, may foster attachment
security along with other aspects of development (Belsky & Cassidy, 1994). Therefore, it has
been argued that future studies need to specify the developmental domains in which early
attachment influences later functioning and to clarify the nature of associations between early
relationships and mastery of later developmental tasks (Belsky & Cassidy, 1994; Thompson,
1999).
It has been argued further that such longitudinal investigations need to include
concurrent representations of relationships, because their associations to adaptive functioning
are stronger in comparison to longitudinal associations, and thus may outweigh the effects of
early attachment on later adjustments (Thompson, 1999). Unfortunately, research on
attachment representations during the transition from childhood to adolescence has not been
abundant (Rice, 1990). The Separation Anxiety Test (SAT; Hansburg, 1980), which in
previous attachment research was modified to be used with young children (Kaplan, 1985;
Klagsbrun & Bowlby, 1976), was originally designed for adolescents in order to identify
separation disorders. Thus, it seemed to be a viable alternative for exploring young
adolescents’ representations of relationships. SAT responses were concurrently associated to
separation reunion responses in early childhood (Shouldice & Steveson-Hinde, 1992; Slough
& Greenberg, 1990) and were longitudinally associated to infancy attachment (Main et al.,
1985). Hence, consistent with Thompson’s (1999) recommendations, the present study used
the SAT as an assessment of relationships’ representations within an investigation of teachers’
ratings of school adaptations and their associations to early attachment relations.

School Competence and Early Attachment
7
The study was carried out with Israeli children whose infant attachment relations to
parents and caregivers were assessed with Ainsworth’ Strange Situation Procedure and who
were raised in kibbutzim that practiced collective sleeping (Sagi et al., 1985). Thus, in
addition to the prospective nature of this study, the present sample was characterized by
several unique attributes. First of all, availability of infant attachment classifications to both
parents allows a prospective evaluation of the relative contribution of relations with different
caregivers to later functioning. Past analyses of developmental adaptations in kindergarten
found that secure infant attachment to the caregiver rather than to parents was the strongest
predictor (Oppenheim, et al., 1988), yet best prediction was achieved by a secure network of
relations (van IJzendoorn, Sagi & Lambermon, 1992). Although different approaches have
been used to evaluate the integrated impact of multiple attachments on later outcomes
(Howes, 1999; van IJzendoorn, et al., 1992), it is also possible that children form qualitatively
different relationships with each caregiver. For example, it has been found that relationships
with mother and father, each influences different domains of functioning in five-years-old
children and neither caregiver played the dominant role in predicting children future
functioning (Verschueren & Marcoen, 1999).
Secondly, the distribution of infant attachment data in this sample was marked by an
overrepresentation of ambivalent classifications and an underrepresentation of avoidant
classifications among insecure infants (Sagi et al., 1985), similar to other Israeli samples and
different from the global distribution (van IJzendoorn & Sagi, 1999). In contrast to avoidant
infants who attempt to minimize attachment needs by expressing interest in the environment,
ambivalent infants have been described as focused on the relationship with the attachment
figure, displaying an exaggerated need for care and attention and minimizing interest in
exploration (Cassidy & Berlin, 1994). Since most studied samples are characterized by an
overrepresentation of avoidant classification among insecure infants, there is very little

School Competence and Early Attachment
8
prospective data about the associations between ambivalent coping with infancy attachment
needs and adaptive strategies for coping with later developmental tasks. Thus, in this study it
may also be possible to explore the long-term adaptive consequences of ambivalent
attachment in infancy.
Finally, collective sleeping represents a considerable deviation from what attachment
theory considers tobe desirable for healthy emotional development (see Aviezer, van
IJzendoorn, Sagi & Schuengel, 1994 for a description and review). Empirical evidence shows
that this child rearing ecology was marked by higher rates of insecure relationships with
mothers but not with fathers (Sagi et al., 1985; Sagi, van IJzendoorn, Aviezer, Donnell &
Mayseless, 1994) as well as by interfered interactive processes between children and mothers
(Aviezer, Sagi, Joels & Ziv, 1999). Yet, over the years since the collection of infancy data
(Sagi, et al., 1985) many of the kibbutzim, but not all of them, abandoned collective sleeping
and instituted home sleeping for children, which increased parental involvement in childcare
and changed family life (Lavi, 1982). Consequently, for some of our participants collective
sleeping was terminated during early childhood, whereas for others the experience of
collective sleeping continued well into their middle childhood and as late as into early
adolescence.
This state of affairs presents a rare opportunity to explore systematically, in an
otherwise rather homogeneous sample, how variations in the caregiving context, i.e., sleeping
arrangement, affect later adaptations. In this respect two competing hypotheses can be
formulated. Based on previous indications that collective sleeping had deleterious effects on
the early relationships of mothers and children, one hypothesis states that compared to longer
exposure to collective sleeping its early termination is associated with an improved school
adjustment. The alternative hypothesis states that adversity associated with collective sleep

School Competence and Early Attachment
9
will be mostly evident in infancy, which is prone to its deleterious effects, but no effects of
sleeping arrangement will be evident later in young adolescence.
The major purpose of this study, therefore, was to evaluate how infancy attachment to
parents contributes uniquely to school adaptive functioning in early adolescence. Concurrent
assessments of IQ, perceived competence and representations of relationships along with
consideration of past life events and the ecology of sleeping arrangement allow for a stricter
test of early attachment’s influence. Thus, it was first hypothesized that infancy attachment to
mother and father will contribute uniquely to the variance in school adaptation of young
adolescent children after considering concurrent representation of relationships, life events
since infancy and children’s sleeping arrangement experience. Secondly, it was hypothesized
that contemporaneous associations will be found in early adolescence between representations
of relationships and school adjustment after considering IQ and self-perceived competence
that in previous research accounted for a significant portion of the predicted variance in
school functioning (Jacobsen et al., 1994; Teo et al., 1996). In addition, given that SAT has
been suggested to be an attachment related measure of representation in early adolescence,
longitudinal associations between SAT responses and early attachment relations were
examined. Finally, two competing hypotheses were evaluated with respect to the effects of
sleeping arrangement on various aspects of school functioning in kibbutz children. One
hypothesis postulated better school adjustment for young adolescents for whom collective
sleep was terminated by early childhood whereas the second hypothesis postulated no effect
for sleeping arrangement in young adolescence.
Method
Participants
Sixty-six children (33 girls and 33 boys) from the Sagi et al., (1985) study were
relocated during early adolescence (mean age 11:10, range = 11:4-12:6) and participated in

School Competence and Early Attachment
10
this study along with their parents and homeroom teachers (n=31, some teachers had in their
classroom more than one participant, range 1-9). Consent was obtained from parents and
children. This sample constitutes 78% of the original sample of 85 infants. Ten families of the
original sample (12%) were out of the country or left no forwarding address, eight families
(9%) refused to participate and one child was seriously ill. Of the children that participated in
the study seven children (11%) no longer lived in a kibbutz, three children moved to a
different kibbutz and the parents of 11 children (17%) were divorced. Thus, this sample is
primarily constituted of kibbutz raised children and does not deviate significantly from the
original sample.
Most of the participants (89%) attended kibbutz schools, which were district schools
for a number of neighboring kibbutzim. Compared to non-kibbutz schools, these schools were
characterized by smaller size classes and less formal relations between teachers, students and
parents, as most teachers were kibbutz members (Liebermann, 1999). Typically in these
schools, learning was conducted in the same classroom (“homeroom”) and one teacher (the
homeroom teacher) taught most subjects and also oversaw children’s progress in those
subjects that were taught by specialists (e.g. math, science). Despite these unique
characteristics kibbutz district schools became more similar to the Israeli school system in
their focusing on academic achievement, use of grades and teaching methods (Liebermann,
1999).
Procedure and Measures.
Data collection was conducted on site, in each kibbutz or hometown, both in infancy
and in early adolescence and involved three sources of information: children, parents and
teachers. Four sets of measures from the larger longitudinal study were used in this
investigation: (a) Infant attachment classifications and relationships’ representations of young
adolescents, (b) parental reports about life events since infancy, (c) young adolescents’ IQ and

School Competence and Early Attachment
11
self reports of perceived competence, and (d) teachers’ reports on young adolescents’ school
functioning. Infancy data were collected in each kibbutz when participants were 13-15 months
old (see Sagi et al., 1985 for further details). Young adolescence data were collected on site in
two visits. In the first visit (a) children’s representation of relationships was assessed, (b) they
were administered a standardized test of intelligence and (c) they responded to Perceived
Competence Scales. In addition parents reported on life events since their child’s infancy. The
second visit took place in participants’ schools at the end of the school year and homeroom
teachers, who were acquainted with the children for at least one year, were interviewed about
each participant.
Measures
Infant attachment classifications. Infant attachment to mother, father and professional
caregiver was observed in the Strange Situation procedure (SSP; Ainsworth, et al., 1978).
Participants were classified into the traditional ABC attachment categories with each of the
three attachment figures (See Sagi et al., 1985 for detailed descriptions). No classifications of
disorganization are available for this sample. Since the particular caregivers that were
responsible for infant care were not involved in the care of these children after age two only
attachment classifications with mother and father were used in this study. Due to technical
problems in the original infancy study, Strange Situation data are missing for several
participants. Thus, with regard to attachment to mother, four were classified as insecure-
avoidant, 35 were classified as secure, and 24 as insecure-resistant (infancy data for three
children were missing). Respective figures for infant-father attachment classifications were 6,
41 and 18 (infancy data for one child were missing). The distribution of attachment
classifications (ABC) with mothers and fathers in the present sample did not differ
significantly from the distributions that were reported in the original sample (Sagi, et
al.,1985), suggesting that participants attrition was unrelated to the quality of attachment

School Competence and Early Attachment
12
relationships. Due to small sample size the avoidant and resistant classifications were
considered as a single insecure group in the analyses.
Separation Anxiety Test (SAT). The SAT (Hansburg, 1980) is a semi-projective test
designed to evaluate individuals’ responses to separation and loss. It is constituted of
pictures depicting separation between a child and a caregiver in different contexts where the
child’s emotional expression is ambiguous. Though originally intended for use with
adolescents, in attachment research the SAT was mostly employed with younger children
and the stimulus pictures were modified accordingly (Klagsbrun & Bowlby, 1976; Shouldice
& Stevenson-Hinde, 1992; Slough & Greenberg, 1990). Given that the present study targeted
young adolescents, some of Hansburg’s (1980) original separation pictures for older children
and adolescents were used. Thus, similar to Kaplan (1987) six pictures depicting mild and
severe separations were chosen from Hansburg’s original set and were presented to all
participants in ascending order of separation distress that was verified in a pilot study.
The mild separation pictures used in this study depict the following situations: 1)
Parents go away for two weeks leaving the adolescent with a sitter. 2) The adolescent is
being transferred to a new school. 3) The family is moving out of the kibbutz (a modification
of the picture “moving to a new neighborhood”). The severe separation pictures depicted the
following situations: 4) The adolescent moves to live with his grandparents for good. 5) The
adolescent is running away from home. 6) A parent is being taken to the hospital in an
ambulance. Similar to Scharf (2001), a modified procedure of administration was adopted
that was first suggested by Klagsbrun and Bowlby (1976) and was later employed by
Kaplan, (1985) and Resnick (1992). Thus, instead of choosing a response from a list of
possible reactions, participants were asked to respond to two open-ended questions that
probed for (a) the feelings of the teenager in the pictured situation (“how feels?), and (b) the
pictured teenager’s actions following the described event (“what does next?”). The latter

School Competence and Early Attachment
13
included attempts to prevent the separation, constructive activities such as getting in touch
with the caregiver, passivity or destructive actions. The entire interview lasted about 20
minutes. It was audiotaped and transcribed verbatim.
Though in previous research attachment classifications were assigned based on SAT
responses (e.g. Shouldice & Stevenson-Hinde, 1992; Slough & Greenberg, 1990), this
approach was more valid when used with younger children as compared with adolescents
(Kroger & Haslett, 1988; Rice, 1990; Scharf, 2001). In order to assess individual differences
of relationships’ representation in the context of imagined separations, three 9-point rating
scales were employed in this study: Coping with Separations, Emotional Openness and
Coherence of transcript. Similar to Scharf (2001) scoring of the two former scales was based
on a system developed by Kaplan (1985) and modified by Resnick (1992) for adolescents,
whereas the latter scale was scored based on the guidelines of Main & Goldwyn (1994).
More specifically, children’s strategies of coping with separation from parents or
familiar people were coded from 1 to 9 according to their adequacy as a solution to the
problems presented in the pictured episodes in terms of a flexible balance between self-
reliance and the ability to turn to others for assistance. Thus, for example active seeking of
information and comfort from alternative close figures is rated as more constructive and
received higher scores than passive solutions, such as not knowing what to do. Lowest levels
of coping involve self-destructive solutions such as self-isolation or even considering self-
hurt. Children’s ability to express openly feelings of vulnerability with regard to separations
and missing of loved ones, such as sadness, loneliness, fear and anger was evaluated from 1
to 9 on Emotional Openness scale. Open and appropriate expressions of vulnerability in the
context of separations from parents earns the adolescent a higher score on this scale because
such expressions indicate that he/she is emotionally affected by the pictured situations
without being overwhelmed. Tense and conflictual responses with regard to such feelings are

School Competence and Early Attachment
14
scored lower as they indicate difficulty with emotional openness about separation and loss.
Finally, the organization, clarity and internal consistency of children’s responses to the
pictured separation episodes were evaluated from 1 to 9 on Coherence of Transcript scale.
Clear, emotionally coherent and appropriately justified responses to the pictures earn the
adolescent a high score on coherence of transcript as opposed to non-relevant responses or
incomplete discussions of the pictured events which received low scores.
A single rater who was blind to the infancy data and uninvolved with other parts of
this study rated SAT transcripts. Another rater independently scored 38% of the data, for
purposes of reliability rating. Using agreement within one-scale-point as matches between
two independent observers, interrater agreement was 88%, 76% and 85% (for Coping,
Emotional Openness and Coherence of Transcript Scales, respectively). In addition,
interrater reliability was assessed by means of intraclass correlations and was .84, .80 and
.89 for Coping, Emotional Openness and coherence of transcript scales, respectively.
However, SAT scales were highly intercorrelated (intraclass r = .95). Thus, in order to avoid
collinearity problems in the analysis, a mean SAT score was computed from the three SAT
scales. Such a score would reflect an overall evaluation of the child’s representations with
regard to separations from close people, similar to categorical classifications used in
assessments such as the Adult Attachment Interview (Main & Goldwyn, 1994).
Life Event Questionnaire (LEQ). The LEQ was administered to the parents in order
to document the participants’ life experiences since infancy. This questionnaire is based on
Life Experiences Survey (Sarason, Johnson, & Siegel, 1978) and on Israel PERI Life Events
Scale (Levav, Krasnoff, & Dohrenwend, 1981) that was adapted for use in Israel by
Eisikovits, Sagi, Guttman, and Sela (1988) and was adjusted to accommodate this sample of
kibbutz raised children. It is a self-report that marks events in life domains such as marital
relationship, parenting and childcare, work, education, loss and trauma as well fluctuations

School Competence and Early Attachment
15
in health, habits and free time activities. The LEQ was separately administered to each
participant’s mother and father who marked occurrence or non-occurrence of each item. In
this context it is important to note that a high degree of homogeneity in socio-economic and
educational status characterized kibbutz population, which was dictated by kibbutz collective
ideology. For example, 77% of kibbutz people had 12-15 years of schooling compared to
54% in the Jewish population at large (Meron, 1994). Thus, socio-economic and educational
statuses were presumed to influence outcome variables less, compared to life events that
were captured by LEQ.
Due to the overinclusive nature of this questionnaire, many of the LEQ items were not
relevant to the questions raised in this study. However, it was thought that those events, which
may influence children’s functioning and representations are likely to involve characteristics
of childcare experience (e.g., insufficient help in childcare, joining a different group of
children) as well as parents’ marital relationships (e.g., parental separation, change in amount
of arguments between the spouses) losses in the family (e.g. a stillbirth, death of a family
member) and changes to parents’ health, capabilities and habits. Correspondingly, those LEQ
items that referred to events in these content domains were aggregated into four separate
clusters: Parenting and childcare, marital relations, trauma and loss, parental transitions.
Based on an assumption that a reported event occurred even if it was only reported by one of
the parents, each reported life event was marked once whether it was reported by one of the
parents or both. Consequently, four life event clusters were constructed for each participant
based on parents’ reports. Sufficiently high alpha reliabilities were obtained only for the
childcare/parenting cluster (11 items, alpha = .67) and the marital relationship cluster (nine
items, alpha = .79). Thus, only these two clusters, which marked (a) childcare quality (i.e.,
number of changes and fluctuations pertinent to childcare arrangements) and (b) quality of
marital relations (i.e., perturbations in the spouse relations) were used in subsequent analyses.

School Competence and Early Attachment
16
Sleeping Arrangement. In addition, we marked for each participant whether the
transition to home sleeping took place before age six (hence home sleeping) or after age six
(hence collective sleeping).
Wechsler Intelligence Test (WISC-R). Participants were tested individually on an
abbreviated version of the WISC-R (Wechsler, 1974) that consisted of two verbal sub-tests
(vocabulary and comprehension) and two performance sub-tests (block design and picture
completion). These sub-tests were selected based on their high correlations to overall IQ
(Liblich, Ben-Shachar-Segev & Ninio, 1976).
Perceived Competence Scales (PCS). Children completed the PCS (Harter, 1982,
1985), which have been translated into Hebrew and were found to replicate the factor
structure delineated by Harter (Assor, Flum & Meir, 1987). Each questionnaire item was
scored from 1 (low perceived competence) to 4 (high-perceived competence). Items’ scores
were averaged to form six orthogonal subscales tapping the following competencies:
Academic, social acceptance, athletic, appearance, behavioral, and general self-worth. Alpha
reliabilities for these scales were .87, .31, .71, .67, .75, .65, respectively. Consequently the
social acceptance scale was omitted from further analyses.
Outcome Measures
Teachers’ ratings of School Functioning. Each participant’s homeroom teacher was
interviewed at school in the end of the school year and reported on children’s achievement
and behavior as well as evaluated it in comparison to an average 6th grade child. Teachers
used a set of rating scales (from very poor to excellent) to evaluate participants’ learning
competence and habits (e.g., task persistence, efficient use of time, verbal skills, curiosity,
motivation), interpersonal relations (e.g. liked by children, behavior with peers) and
disciplinary problems (e.g. Interruptive behavior, need for disciplinary measures).
Consequently, two dimensions were identified: Academic functioning (9 items) and social-

School Competence and Early Attachment
17
emotional functioning (10 items). Items were standardized and were submitted to two separate
exploratory factor analyses to confirm this structure of the data (Tinsley & Tinsley, 1987),
one for each dimension. Analysis 1 tested the loading of academic functioning items on two
hypothesized factors: Scholastic skills and scholastic attitude, whereas analysis 2 tested the
loading of social-emotional functioning items on two hypothesized dimensions: Social
competence and behavioral difficulties. Using a principal component factor extraction with
varimax rotation, analysis 1 extracted two factors for academic functioning and analysis 2
extracted three factors for social emotional functioning. Thus, overall the hypothetical factor
structure was confirmed. Following this analysis items’ scores were summed to create the
following dependent measures of school adaptation: Scholastic Attitude, Scholastic Skill,
Social Competence, Emotional Maturity and Behavioral Difficulties. Table 1 presents factor
loading and alpha reliabilities for each factor as well as percentage of explained variance and
Eigenvalues, separately for each analysis. Two additional dependent measures for each
participant were Grade Point Average (GPA), which were obtained from the schools as well
as teachers’ overall evaluation of participants’ academic progress on a six-point scale from
poor to excellent, compared to a 6th grade child.
Insert Table 1 about Here
Statistical Analyses.
In the first step, correlational analyses were performed to determine the relations of
predictor variables to measures of school functioning. In the second step, a series of
simultaneous multiple regressions was conducted in which school functioning measures were
simultaneously regressed on predictor variables to assess their relative contribution to the
variance in outcome measures. In the third step, hierarchical multiple regression analyses
were performed, in which infancy attachment was entered last in order to assess its unique
contribution to the prediction of young adolescents’ adaptive school functioning over and

School Competence and Early Attachment
18
above the contribution of control variables, representation of relationship in young
adolescence and past life events.
Results
Analysis Step 1: Intercorrelations between predictor variables and outcome measures
The correlations between predictor variables and outcome variables are presented in
Table 2. As expected, moderately high positive correlations were found between Perceived
Academic Competence as well as IQ to all outcome measures that are academically oriented
and to Emotional Maturity. In addition, teachers’ ratings of Behavioral Difficulties were
negatively correlated to participants’ Perceived Academic Competence and were unrelated to
their IQ. Furthermore, teachers’ rating of Behavioral Difficulties showed a moderately high
negative correlation to Perceived Behavioral Competence. However, Perceived Appearance,
Athletic Competence and General self-worth were not correlated to any outcome measure and
were therefore omitted from further analyses. It is important to note that the associations
between measures of school adaptations and scales of perceived competence probably reflect
interdependent relations between self-concept and adaptive functioning. However, in order to
expose the unique contribution of infancy attachment to young adolescents’ school
adaptation, it was necessary to control for the variance explained by relations to concurrently
assessed variables. Similar to Perceived Academic Competence, SAT showed positive, low to
moderate correlation to all measures of school adaptation except for Behavioral Difficulties.
Thus, representations of relationships in early adolescence as marked by SAT were correlated
to academic as well as socioemotional outcome variables. It is interesting to note that
teachers’ rated Social Competence in this sample was unrelated to both IQ and Perceived
Academic Competence.
Insert table 2 about here

School Competence and Early Attachment
19
Furthermore, a moderate significant correlation was found between infancy
attachment to mother and Scholastic Skills and a moderate, but only marginally significant
correlation was found between infancy attachment to mother and Emotional Maturity.
Contrary to expectations infancy attachment to father was not correlated to any outcome
measure. With regards to the relations between preceding life events and school adaptation,
Table 2 shows a significant negative correlation between perturbations in spouse relations and
Scholastic Attitude, as well as marginally significant negative correlations between this life
event variable and Emotional Maturity and Social Competence. Unexpectedly, fluctuations in
childcare were not correlated to any measure of school adaptation, thus LEQ childcare cluster
was omitted from further analyses. However, a significant correlation was found between
sleeping arrangement and Emotional Maturity. This correlation suggests that participants who
moved to home sleeping by early childhood were more emotionally mature. Finally, Table 2
shows that children’s gender was not correlated to any outcome measures, thus it was not
entered into further analyses.
Before conducting data analysis the intercorrelations of predictor variables were
examined. No significant correlations were found for infancy attachment to mother as well as
SAT scores with the other predictor variables (life events’ cluster, sleeping arrangement, self-
perceived competence scales and IQ). The correlation between infancy attachment to mother
and SAT score was only marginally significant (r= - .21, p < .10) indicating a tendency for
discontinuity between early attachment classifications and later ratings of relationships’
representations. Also, even though SAT was not correlated with overall IQ the question may
be raised about possible confounding of SAT scores with verbal skills. The Hebrew WISC-R
manual (Liblich, et al., 1976) shows that Vocabulary is the verbal sub-test with the best
psychometric qualities and highest intercorrelations with verbal IQ. Yet, no significant
correlations were found between Vocabulary score and SAT scales’ scores, thus suggesting

School Competence and Early Attachment
20
that SAT scores were not confounded with verbal IQ. Finally, in our data self-perceived
academic competence and self-perceived conduct competence were moderately correlated (r =
.41, p < .01). However, in the absence of consensus about the size of intercorrelations among
predictor variables that create multicollinearity (Pedhazur, 1982), these two variables were
entered to the analyses.
Analysis Step 2: Simultaneous multiple Regressions of School Adaptation Measures on
Predictor variables
In order to determine that infancy attachment relations and sleeping arrangement
exerted unique effects on outcome measures, it was necessary to establish that these variables
did not interact with any other predictor variables with regard to the outcome measures.
Therefore, initial analyses computed a series of linear regressions for all school adaptation
measures by entering infancy attachment to mother and one other predictor variable (IQ,
Perceived Competence, SAT, LEQ clusters and sleeping arrangement) together with their
interaction term. A similar procedure was carried out for sleeping arrangement variable and
infancy attachment to father. In addition, it was established that none of the dichotomous
predictor variables (infancy attachment to mother, infancy attachment to father and sleeping
arrangement) interacted with regard to the outcome measures by means of a set of 2 (secure,
insecure) X 2 (collective, home) two-way ANOVAs as well as a set of 2 (secure, insecure) X
2 (secure, insecure) two-way ANOVAs. These initial analyses yielded only one interaction
effect for the prediction of Social Competence in which infant attachment security to mother
interacted with LEQ cluster of spouse relations. In this interaction effect, high perturbations in
parents’ spouse relations were related to lower Social Competence score primarily for
adolescents who were insecurely attached to mother in infancy. Therefore, this interaction
effect was incorporated into the regression equation of Social Competence. The absence of
interaction effects between infancy attachment to mother and infancy attachment to father

School Competence and Early Attachment
21
indicates no compensation effects for insecurity in either relationship. Consequently infancy
attachment to father was omitted from further analyses.
Insert Table 3 about here
Table 3 summarizes the simultaneous multiple regression analyses of school
adaptation measures on predictor variables. The regression equations that were obtained for
school adaptation measures were all significant, but hypothesized predictor variables (infancy
attachment to mother, SAT score or Sleep arrangement) contributed significant portions of the
variance only in some outcome measures. Thus, the regression equation for Emotional
Maturity was significant with multiple R = .72, R2 = .52 (F [7,52] = 7.95, p< .000), and the
predicted variance was accounted for by infancy attachment to mother, SAT and Sleeping
Arrangement in addition to controlling for IQ and Perceived Academic Competence.
Similarly, the regression equation for Scholastic Skills was significant with multiple R = .80,
R2 = .64 (F [5,52] =19.07, p< .000) and both infancy attachment to mother as well as SAT
accounted for significant portions of the variance in addition to controlling for IQ and
Perceived Academic Competence. The regression of Social Competence on the linear
combination of SAT, LEQ spouse relation cluster and an interaction term of LEQ Spouse
relation cluster and infancy attachment to mother was also significant with multiple R = .44,
R2 = .19 (F [3,57] = 4.43, p< .007). Both SAT and the interaction term of infancy attachment
to mother by LEQ spouse relations contributed significant portions of the variance in this
outcome measure.
For General Progress, multiple R = .73, R2 = .54 (F [3,58] = 22.48, p< .000), while for
GPA multiple R = .64, R 2= .44 (F [3,53] = 12.43, p < .000). In both regression equations
SAT contributed a significant portion of the variance in addition to Perceived Academic
Competence. It is interesting that IQ contributed significantly to the predicted variance in
General Progress, but not in GPA. Finally, though the regression equation of Scholastic

School Competence and Early Attachment
22
Attitude was significant with multiple R = .58, R2 = .55 (F [6,51] = 4.20, p< .002), the
variance in this measure was primarily accounted for by Perceived Academic Competence.
The contributions of infancy attachment to mother and SAT to the variance in Scholastic
Attitude were short of significance. Similarly, though the regression equation of Behavior
Difficulties was significant with multiple R = .52, R2 = .27 (F [4,58] = 5.32, p< .001) the
variance in this variable was primarily accounted for by Perceived Behavioral Competence.
Analysis Step 3: Hierarchical Multiple Regression of School Adaptation Measures on
Predictor Variables Controlling for IQ and Perceived Competence.
For those outcome measures in which simultaneous regression equations showed that
infancy attachment, SAT and sleeping arrangement contributed significantly to the predicted
variance, hierarchical multiple regression analyses were conducted in order to determine that
these predictor variables contributed uniquely to outcome measures, beyond and over the
contribution of IQ and Perceived Competence. Accordingly, the model of hierarchical
multiple regressions followed a chronological rule in which variables to be controlled for and
concurrently assessed with outcome measures were entered earlier, whereas variables which
their unique contribution needed to be determined and happened earlier in participants’ life
were entered later. Thus, IQ and Perceived competence scales were entered first in one block.
SAT was entered second because it was assessed concurrently with outcome measures and
was, thus, likely to have a major contribution to the predicted variance in outcome measures.
Past life events (LEQ spouse relations cluster) and sleeping arrangement, which marked life
events that predated assessment of school functioning and thus may mediate the influence of
early attachment were entered together in the third block. Infancy attachment to mother was
entered last in order to examine whether it had a unique contribution to the predicted variance
in school adaptation, beyond and above all other predictor variables.

School Competence and Early Attachment
23
Table 4 summarizes the results of the hierarchical multiple regression analyses, which
showed that SAT contributed significantly to the prediction of Scholastic Skill (R2 change =
.15, F[1,54] = 19.47, p< .000), General Progress (R2change = .04, F[1,58] = 4.48, p< .04) and
GPA (R2 change = .08, F[1,52] = 7.11, p< .01) over and above IQ and Perceived Academic
Competence. However, only a marginally significant unique contribution of SAT was found
for Scholastic Attitude (R2 change = .045, F[1,55] = 3.54, p< .07). Unexpectedly, SAT had no
unique contribution to Emotional Maturity. Sleeping Arrangement had a unique contribution,
over and above control variables and SAT, only to Emotional Maturity (R2 change = .10, F[1,
55] =10.64 , p< .002). Finally, when entered last after most of the predicted variance was
accounted for, infancy attachment to mother still added a significant unique portion to the
predicted variance of young adolescents’ Scholastic skills (R2change = 08, F[1,53] = 11.48,
p< .001) and Emotional Maturity (R2change = .05, F[1,54] = 5.22, p< .026).
Since Social Competence was not interrelated with control variables SAT was entered
first in the regression equation of this variable, the interaction term of infant attachment by
LEQ spouse relation cluster was entered next, and LEQ spouse relation cluster was entered
last. Thus, it is not surprising that SAT, which was entered in the first block, contributed
significantly to the predicted variance in Social Competence (R2 change = .06, F[1, 59] =
4.03, p< .05). The interaction term of infancy attachment by LEQ spouse relation did not
contribute significantly to the predicted variance of Social Competence, however LEQ spouse
relation cluster contributed significantly to the variance of Social Competence (R2 change =
.11, F[1,57] = 7.89, p< .007) when entered last after SAT’s contribution was accounted for.
Thus, it can be summarized that SAT contributed uniquely to Scholastic Skills,
General Progress and GPA and showed a trend in Scholastic Attitude, over and beyond IQ
and Perceived Academic Competence. Sleeping Arrangement contributed uniquely to the
predicted variance of Emotional Maturity, over and beyond IQ and Perceived Academic

School Competence and Early Attachment
24
Competence, while LEQ spouse relations cluster contributed uniquely to the variance in
Social Competence. Furthermore, infancy attachment to mother contributed uniquely to the
predicted variance of Scholastic Skills and Emotional Maturity, over and beyond the
contribution of all predictor variables.
Finally, to clarify the relations of the findings to early ambivalent attachment relations
rather than to early attachment insecurity, step 3 analyses were repeated without the
participants who were classified as avoidant in infancy. Despite considerable reduction in
sample size the overall pattern of findings was preserved even though some of the effects
were reduced. Thus, SAT (R2change = .11 F[1,43] = 10.91, p<.002) and infancy attachment to
mother (R2change = .08, F[1,42] = 9.25, p< 004) contributed uniquely to the variance in
Scholastic Skills (n=47), over and beyond the influence of IQ and Perceived Academic
Competence. Similarly, The unique contribution of SAT to the predicted variance in General
Progress (n=49; R2change = .05, F[1,46] = 5.13, p< .03) and GPA (n=44; R2change = .06,
F[1,41] = 4.48, p< .04), over and above the contribution of control variables was preserved.
With regards to the variance in Emotional Maturity (n=49) it was still significantly predicted
by Sleeping arrangement (R2change = .08, F[1,46] = 5.13, [1,43], p< .05), however the
contribution of infancy attachment was reduced to a trend (R2change = .035, F[1,42] = 2.91,
p<.095).
Discussion
Results of this study showed that school adaptation in early adolescence was
concurrently linked to representations of relationships as well as longitudinally associated to
infancy attachment and to changes in the context of caregiving, beyond concurrently assessed
IQ and self-perceived competence. Although self-perceived competence and sometimes IQ,
when entered first, contributed the largest amount of explained variance in all the dimensions
of school adaptation except social competence, significant portions of the variance were still

School Competence and Early Attachment
25
explained by representations of relationships and infancy attachment security. As expected
with regard to contemporaneous connections, young adolescents whose representations of
coping with separation and loss were coherent, emotionally regulated and constructive were
also rated by their teachers as competent in most academic domain variables of school
functioning, beyond IQ and self-perceived competence. Finding that representations of
relations were significantly associated to GPA, general academic progress and scholastic
skills is congruent with previous research (Jacobsen & Hoffman, 1997), and it indicates that
such representations of relationships may support children’s fuller engagement with academic
tasks.
The large effects of self-perceived academic competence on school adaptation are
congruent with the notion that self-perceived competence and school adaptation may be part
of feedback loops between self-perceived control and actual outcome that constitute self-
confirming cycles (Skinner, Zimmer-Gembeck & Connell, 1998). Furthermore, according to
this motivational model, underlying children’s perception of control are two dimensions of
the social context; structure and involvement that together resemble the sensitivity construct
in attachment research (Skinner, et al., 1998). Thus, it may be argued that in challenging
interpersonal situations of coping with separations and loss a sense of higher control is
reflected in constructive, coherent and emotionally regulated representations, which allows
children to deal with academic tasks more effectively.
With regard to longitudinal connections between early adaptations defined in terms of
infancy attachment security that was entered last into the regression analyses and later school
adjustment, the proposed predictive associations have been partly confirmed. Young
adolescents who were securely attached to mother in infancy received higher scores on
Scholastic Skills, which were evident in higher teachers’ ratings of curiosity and verbal
abilities, beyond the influence of control variables and concurrent representations of

School Competence and Early Attachment
26
relationships. In addition, these adolescents received higher scores on Emotional Maturity,
which were manifested in higher ratings of ability to work independently and cope with
frustrations and criticisms, beyond the influence of control variables and changes in
caregiving circumstances.
These findings are concordant with previous research that found predictive links from
secure representations to cognitive functioning, GPA, and school behavior (Jacobsen et al.,
1994; Jacobsen & Hoffman, 1997). In addition, these results along with the findings of Teo et
al., (1996) underscore the importance of considering adaptations in infancy for predictions of
later adjustments in addition to more current circumstances. Moreover, finding that early
attachment to mother contributed significantly to predictions of Scholastic Skills and
Emotional Maturity after concurrently assessed, more proximal, variables have been entered
implies that in predicting aspects of school adaptation early experience still adds to later
assessments (Sroufe, Carlson, Levy & Egeland, 1999). More Specifically, Scholastic Skills
and Emotional Maturity may be viewed as dimensions of school adaptation that reflect
exploration in its broad sense that includes organized behavior, interest and motivation
(Ainsworth et al., 1978).
The links of school adjustment in early adolescence to concurrent representations of
relationships as well as to attachment security in infancy are congruent with the notion of a
secure-base control system (Ainsworth, et al., 1978; Bowlby, 1988). Thus, children’s
confidence in a relationship allows them to explore away from secure base and master
successfully ordinary environments such as schools (Waters & Cummings, 2000).
Accordingly, variations in adaptive school functioning in early adolescence may mirror the
secure-base control systems children had set up with their mother based on their attachment
experiences during infancy. However, as underscored by Waters and Cummings (2000),
because the establishment of set-goals for the secure base system is unlikely to be finalized in

School Competence and Early Attachment
27
infancy without further learning and modifications, it is necessary to consider the continued
influence of caregiving.
In this study caregiving was only indirectly addressed by considering the ecological
context of childcare in terms of collective versus home sleeping arrangement and the quality
of marital relations. With regard to sleeping arrangement two competing hypotheses were
considered. The data did not support the hypothesis that deleterious effects of collective
sleeping were restricted to infancy. In this study, young adolescents who were exposed to
collective sleeping beyond early childhood were rated as less emotionally mature in the
school context compared with young adolescents who slept with parents. Thus, though not all
encompassing of all school adaptation domains, the adversity associated with collective
sleeping appeared to be cumulative and particularly influential on young adolescents’ ways of
managing their schoolwork. These finding are concordant with suggestions that changes in
caregiving context may alter children’s experience with parents and consequently influence
their functioning and internal representations of relationships (Lewis, Feiring & Rosenthal,
2000; Weinfield, Sroufe & Egeland, 2000).
The change in sleeping arrangement from collective to home sleeping may also
underlie the absence of associations between early attachment security and SAT
representations. It is possible that the interference to dyadic interactive processes suggested
for the collective sleeping ecology (Aviezer et al., 1999) was alleviated with the change to
home sleeping. Furthermore, even though no interaction effect of sleeping arrangement by
SAT was revealed, finding long-term predictive associations between earlier home sleeping
and later emotional maturity in the school context suggests that children’s relations with
parents may have been experienced differently in the two sleeping arrangements ecologies.
In a similar vein, finding predictive association between high perturbations in parents’
marital relations and later lower social competence indicates a possible link between

School Competence and Early Attachment
28
satisfactory marital relationships and aspects of caregiving quality that manifested themselves
in children’s social competence at school. This finding is concordant with evidence that
linked satisfactory marital relationships to supportive caregiving and secure attachment
(Belsky, 1999). Thus, different aspects of the child-rearing context may exert long-term
effects on developmental outcomes by influencing the parent-child relations and children’s
experiences with their parents.
Previous research marked close relationships as the main, theoretically coherent,
outcome domain that is influenced by early attachment experiences (Thompson, 1999;
Weinfield, et al., 1999). However, the results of this study suggest additional outcome
domains that are longitudinally correlated with infant attachment security. According to
Cassidy and Berlin (1994) ambivalently attached individuals are characterized by constrained
autonomy and restricted exploratory competence. In this study ambivalence was
overrepresented among infant insecure classifications and the results were largely maintained
even when analyses compared only youngsters with ambivalent and secure attachment to
mother in infancy. Thus, it may be speculated that the significantly lower ratings of scholastic
skills and emotional maturity that were found in young adolescents who as infants were
insecure with mothers are associated with ambivalence in infancy.
Although several studies found lower cognitive and metacognitive functioning among
insecure individuals who were predominantly avoidant (e.g. Frankel & Bates, 1990;
Lutkenhaus, Grossmann, Grossmann, 1985; Main, 1991; Matas, et al., 1978; Moss, et al.,
1997), most of these studies were conducted with younger children. Previous studies with
older children showed only few associations between early attachment security and later
curiosity and autonomous approach to tasks. This may reflect the difficulty to distinguish
exploratory behavior of secure individuals who are driven by a genuine sense of autonomy
from exploratory behavior of avoidant individuals who are driven by defensive maneuvers

School Competence and Early Attachment
29
against attachment needs. On the other hand, ambivalent individuals’ dependency tendencies,
exaggerated attachment needs and decreased explorations are likely to stand out in
comparison to the autonomy and readiness to explore away from attachment figures that
characterize secure individuals. Therefore, our finding of lower scholastic skills and
emotional maturity for adolescents whose infancy attachment to mother was insecure reflect
the overrepresentation of ambivalence among insecure infants in our sample and support the
discriminant validity of early attachment patterns. In addition, these findings support Sroufe,
et al., (1999) arguments for the special role of early experience as paving specific pathways of
long-term developmental progression and adaptations.
Children’s rated social competence and behavior difficulties at school were not
significantly linked to early attachment security and were only marginally influenced by
contemporary representation of relationships. This was unexpected in view of prior theory and
research that suggested an early attachment influence in the domain of social behavior and
peer relationships (Thompson, 1999). The lack of associations between early attachment and
teachers’ rated social competence may have resulted from teachers’ focusing on general
tendencies of behavior in social contexts rather than on close personal relationships. It has
been argued that teachers’ ratings of social relations with peers may not be linked closely to
child parent attachment (Schneider, Atkinson & Tardif, 2001), as teachers may be more able
raters of students’ capabilities in school-related tasks and less skillful in observing students’
competence in peer relations. In addition, previous research concerned with association of
attachment relations to behavioral problems has been inconclusive possibly because in normal
samples behavioral difficulties are determined by multiple factors (Thompson, 1999).
Jacobsen and Hofmann (1997) reported similar finding with regards to teachers’ ratings of
social behavior and behavior disruptions. Our findings along with their findings suggest that
teacher rating may by unsuitable as an evaluation of these behavioral characteristics of

School Competence and Early Attachment
30
adjustment to school. In addition, further research is needed to determine whether early
attachment relations interact with specific age related tasks, such as adjustment to the
transition into adolescence, to produce behavioral difficulties in some children.
Contrary to expectations, early attachment to father did not have any predictive
associations to school functioning in early adolescence. The absence of longitudinal
associations between infant father attachment and later adjustments resembles Freitag et al.,
(1996) findings in which infant father attachment did not contribute to later adaptations in the
domain of peer relations. These findings are perplexing because fathers’ sensitivity during
play with toddlers was associated with less avoiding styles of coping during adolescence
(Zimmermann & Grossmann, 1997). In addition, competent functioning at school may be
viewed as epitomizing the exploration side of the attachment-exploration balance in which
fathers have been postulated to be particularly supportive (Grossmann, et al., 1999). However,
it is important to note that even though the participants were extensively exposed to other
caregivers, mothers were considered as the primary caregivers in kibbutzim and spent more
time with children than fathers (Aviezer, et al., 1994). Thus, children may have experienced
their relations with each parent differently and our findings suggest that the dynamics of
attachment relationship with mothers and fathers may be associated with different
longitudinal meanings. Given the dearth of longitudinal research that had examined the
developmental consequences of infant father attachment relations it is suggested that further
research is needed to focus specifically on attachment relations in infancy.
Before concluding it should be noted that in the absence of reliable and valid
assessments of attachment representations in young adolescence an adaptation of Kaplan’s
(1987) scoring scheme of the SAT was used in this study. However, given that this semi-
projective measure was not originally devised to assess attachment representations and that
the available scoring schemes have been geared to young children, it may require further

School Competence and Early Attachment
31
validation to become an assessment of attachment internal representations that is suitable for
young adolescent children. Despite these reservations the findings indicate some overall
continuity from infancy to early adolescence in established set-goals for the secure base
system with mothers that was not overshadowed by changes in contextual circumstances or
contemporaneous representations of relationships. In particular this study indicates that
coping with later developmental tasks epitomizes earlier strategies that may have been
developed for coping with infancy attachment needs.
Finally, it is necessary to keep in mind that this study was conducted with a unique
sample with specific characteristics, namely kibbutz children who come from a unique
ecological background. In addition this sample is unique in the prevalence of ambivalence
among infancy insecure classifications. Thus, caution is required when the findings are
examined and generalized. Nonetheless, considering the 10-year time lag between
assessments in this study and complete independence of the measurements used in infancy
and young adolescence, these results present an intriguing picture of longitudinal associations
that imply positive influences from infant-mother attachment security to later school
competence.

School Competence and Early Attachment
32
References
Ainsworth, M.D.S., Blehar, M., Waters, E. & Wall, S. (1978). Patterns of attachment:
A psychological study of the strange situation. Hillsdale, NJ: Erlbaum.
Assor, A., Flum, H., & Meir, Y. (1987). Overrating and underrating of academic
competence in children. Paper presented at the annual meeting of the Israeli Psychological
Association, Tel-Aviv.
Aviezer, O., Sagi, A. Joels, T. Ziv, Y. (1999). Emotional Availability and attachment
representations in kibbutz infants and their mothers. Developmental Psychology, 35, 811-821.
Aviezer, O. van IJzendoorn, M. H., Sagi, A. & Schuengel, C. (1994).”Children of the
Dream” revisited: 70 years of collective child care in Israeli kibbutzim. Psychological
Bulletin, 116, 99-116.
Belsky, J. (1999). Interactional and contextual determinants of attachment security. In:
J. Cassidy & P. R. Shaver, (Eds.), Handbook of attachment: Theory, research and clinical
applications (pp. 249-264). New York: The Guilford Press.
Belsky, J, & Cassidy, J. (1994). Attachment: Theory and evidence. In M. Rutter, & D.
Hay, (Eds.), Development through life: A Handbook for clinicians (pp. 373-402). Oxford:
Blackwell.
Bowlby, J. (1988). A Secure Base. Basic Books.
Bretherton, I. (1997). Bowlby’s legacy to developmental psychology. Child
Psychiatry and Human Development, 28, 33-43.
Cassidy, J. (1986). The ability to negotiate the environment: An aspect of infant
competence as related to quality of attachment. Child Development, 57, 331-337.
Cassidy, J., & Berlin, L. (1994). The insecure/ambivalent pattern of attachment:
Theory and research. Child Development, 65, 971-981.

School Competence and Early Attachment
33
Dornbusch, S. M., Ritter, P. L., Leiderman, P. H., Roberts, D.F., & Fraleigh, M. J.
(1987). The Relation of parenting style to adolescent school performance. Child Development,
58, 1244-1257.
Eccles, J. S., Lord, S., & Buchanan, C. M. (1996). School transitions in early
adolescence: What are we doing to our young people? In J. A. Graber, J. Brooks-Gunn & A.
Petersen (Eds.), Transitions through adolescence (pp. 251-284). Mahwah, NJ: Lawrence
Erlbaum Associates, Publishers.
Eisikovits, Z., Sagi, A., Guttmann, E., & Sela, M. (1988). Wife abuse in Israel: An
ecological perspective. Paper presented at the Third Conference on Developmental
Psychology, Budapest, Hungary.
Frankel, K., & Bates, J. (1990). Mother-toddler problem solving: Antecedents in
attachment, home behavior, and temperament. Child Development, 61, 810-819.
Freitag, M.K., Belsky, J., Grossmann, K., Grossmann, K.E., & Scheuerer-Englisch, H.
(1996). Continuity in parent-child relationships from infancy to middle childhood and
relations with friendship competence. Child Development, 67, 1437-1454.
Frodi, A., Bridges, L., & Grolnick, W. (1985). Correlates of mastery related behavior:
A short-term longitudinal study of infants in their second year. Child Development, 56, 1291-
1299.
Grossmann, K. E., Grossmann, K., & Zimmermann, P. (1999). A wider view of
attachment and exploration: Stability and change during the years of immaturity. In: J.
Cassidy & P. R. Shaver, (Eds.), Handbook of attachment: Theory, research and clinical
applications (pp. 760-786). New York: The Guilford Press.
Hansburg, H. G. (1980). Adolescent Separation Anxiety: A Method for the study of
adolescent separation problems. Huntington, New York: Robert E. Krieger Publishing
Company

School Competence and Early Attachment
34
Harter, S. (1982). The perceived competence scale for children. Child Development,
53, 87-97.
Harter, S. (1985). Competence as a dimension of self-evaluation toward a
comprehensive model of self-worth. In: R. L. Leahy (Ed.) The development of the self (pp. 55-
121). New York: Academic Press, inc.
Howes, C. (1999). Attachment relationships in the context of multiple caregivers. In:
J. Cassidy & P. R. Shaver, (Eds.), Handbook of attachment: Theory, research and clinical
applications (pp. 671-687). New York: The Guilford Press.
Hazen, L. N., & Durrett, M. E. (1982). Relationship of security of attachment to
exploration and cognitive mapping abilities in 2-year olds. Developmental Psychology, 18,
751-759.
Jacobsen, T., Edelstein, W., & Hofmann, V. (1994). A longitudinal study of the
relation between representations of attachment in childhood and cognitive functioning in
childhood and adolescence. Developmental Psychology, 30, 112-124.
Jacobsen, T., & Hofmann, V. (1997). Children’s attachment representations:
longitudinal relations to school behavior and academic competency in middle childhood and
adolescence. Developmental Psychology, 33, 703-710.
Kaplan, N. (1985). Procedures for administration of the Hansburg separation Anxiety
Test for younger children. Unpublished manuscript, Department of Psychology, University Of
California, Berkeley.
Klagsbrun, M., & Bowlby, J. (1976). Responses to separation from parents: A clinical
test for young children. British Journal of Projective Psychology, 21, 7-27.
Kroger, J., & Haslett, S. J. (1988). Separation-Individuation and ego identity status in
late adolescence: A two-year longitudinal study. Journal of Youth and Adolescence, 17, 59-
79.

School Competence and Early Attachment
35
Lavi, Z (1982). Correlates of sleeping arrangements for infants in kibbutzim. Paper
presented at the International Conference of Infant Studies, New York.
Learner, D.G., & Kruger, L. J. (1997). Attachment, self-concept, and academic
motivation in high-school students. American Journal of Orthopsychiatry, 67, 485-492.
Levav, I., Krasnoff, L., & Dohrenwend, B. S. (1981). Israeli PERI Live Event Scale:
Rating of events by a community sample. Israeli Journal of Medical Sciences, 17, 176-183.
Lewis, M., Feiring, C., & Rosenthal, S. (2000). Attachment over time. Child
Development, 71, 707-720.
Liblich, A. Ben Shachar-Segev, N., & Ninio, A. (1976). WISC-R; Wechsler’s
Intelligence Test for Children. Jerusalem: Institute for developmental Psychology, The
Hebrew University, School Counseling Services, Ministry of Education. (Hebrew).
Liebermann, Y. (1999). Elementary school: From ideological commitment to
professionalism. In: W. Folling & M. Folling-Albers (Eds.) The Transformation of Collective
Education in the Kibbutz (96-105). Frankfurt am Main, Germany: Peter Lang.
Lutkenhaus, P., Grossmann, K.E., Grossmann, K. (1985). Infant-mother attachment at
twelve months and style of interaction with a stranger at the age of three years. Child
Development, 56, 1538-1542.
Main, M. (1991). Metacognitive knowledge, metacognitive monitoring and singular
(coherent) versus multiple (incoherent) models of attachment: findings and directions for
future research. In: C. M. Parks, J. Stevenson-Hinde & P. Maris (Eds.). Attachment across the
life cycle (pp. 127-159). London: Routledge.
Main, M. & Goldwyn, R. (1994). Adult attachment rating and classification systems.
Version 6. Unpublished manuscript, University of California, Berkeley, Department of
Psychology.

School Competence and Early Attachment
36
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood and
adulthood: A move to the level of representation. In: I. Bretherton & E. Waters (Eds.),
Growing points of attachment theory and research. Monographs of the Society for Research in
Child Development, 50(1-2,serial no. 209), 66-104.
Masten, A.S., Coatsworth, J. D., Neemann, J., Gest, S.D., Tellegen, A., & Garmezy,
N. (1995). The structure and coherence of competence from childhood through adolescence.
Child Development, 66, 1635-1659.
Matas, L., Arend, R., & Sroufe, L. A. (1978). Continuity of adaptation in the second
year: the relationship between quality of attachment and later competence. Child
Development, 49, 547-556.
Meron, S. (1994). The Kibbutz movement in 1993: A statistical annual book. Efal,
Israel: Yad Tabenkin (Hebrew)
Moss, E., Gosselin, C., Parent, S., Rousseau, D., & Dumont, M. (1997). Attachment
and joint problem-solving experiences during the preschool period. Social Development,
Moss, E., Parent, S., Gosselin, C., & Dumont, M. (1993). Attachment and the
development of metacognitive and collaborative strategies. International Journal of
Educational Research, 19, 555-571.
Oppenheim, D., Sagi, A., & Lamb, M. E. (1988). Infant adult attachments on the
kibbutz and their relation to socioemotional development four years later. Developmental
Psychology, 24, 427-433.
Pappini, R. D., & Roggman, L. A. (1992). Adolescent perceived attachment to parents
in relation to competence, depression, and anxiety. Journal of Early Adolescence, 12, 420-
439.
Pedhazur, E.J. (1982). Multiple Regression in Behavioral Research (2nd edition). New
York: Holt Reinhart and Winston.

School Competence and Early Attachment
37
Resnick, G. (1992). Measuring attachment in early adolescence: A manual for the
administration, coding and interpretation of the Separation Anxiety Test for 11 to 14 years
old. Unpublished manuscript.
Rice, K. G. (1990). Attachment in adolescence: A narrative and meta-analytic review.
Journal of Youth and Adolescence, 19, 511-538.
Roisman, G. I., Bahadur, M. A, & Oster, H. (2000). Infant attachment security as a
discriminant predictor of career development in late adolescence. Journal of Adolescent
Research, 15, 531-545.
Sagi, A., lamb, M. E., Lewkowicz, K. S., Shoham, R., Dvir, R., & Estes, D. (1985).
Security of infant-mother,-father, and –metapelet attachment among kibbutz-reared Israeli
children. In I. Bretherton & E. Waters (Eds.), Growing points of attachment theory and
research. Monographs of the Society for Research in Child Development, 50(1-2,serial no.
209), 257-275.
Sagi, A., van IJzendoorn, M. H., Aviezer, O., Donnell, F., & Mayseless, O. (1994).
Sleeping out of home in a kibbutz communal arrangement: It makes a difference for infant-
mother attachment. Child Development, 65, 992-1004.
Sarason, I.G., Johnson, J. H., & Siegel, J. M. (1978). Assessing the impact of life
change: Development of the life experiences survey. Journal of Consulting and Clinical
Psychology, 46, 932-946.
Scharf, M. (2001). A “natural experiment” in childrearing ecologies and adolescents’
attachment and separation representations. Child Development, 72, 236-251.
Schneider, B. H., Atkinson, L., & Tardif, C. (2001). Child-parent attachment and
children’s peer relations: A quantitative Review. Developmental Psychology, 37, 86-100.

School Competence and Early Attachment
38
Shouldice A., & Stevenson-Hinde, J. (1992). Coping with security distress: The
Separation anxiety test and attachment classification at 4.5 years. Journal of Child Psychology
and Psychiatry, 33, 331-348.
Skinner, E.A., Zimmer-Gembeck, M.J., & Connell, J.P. (1998). Individual differences
and the development of perceived control. Monographs of the Society for Research in Child
Development, 63(2-3, serial no. 254), 257-275
Slough, N. M., & Greenberg, M. T. (1990). Five-year-olds’ representations of
separation from parents: Responses from the perspective of self and other. In: I. Bretherton
and M. W. Watson, (Eds.) Children’s Perspective on the Family. New Directions for Child
Development, 48, 67-84. San Francisco, Jossey-Bass.
Sroufe, L. A., Carlson, E., Levy, A. K. & Egeland, B. (1999). Implications of
attachment theory for developmental psychopathology. Development and Psychopathology,
11, 1-13.
Sroufe, L. A., Carlson, E., & Shulman, S. (1993). Individuals in relationships:
Development from infancy through adolescence: In D. C. Funder, R. Parke, C. Tomlinson-
Keesey, & K. Widaman (Eds.), Studying lives through time: Approaches to personality and
development (pp. 315-342). Washington DC: American Psychological Association.
Sroufe, L. A., Egeland, B., & Kreutzer, T. (1990). The fate of early experience
following developmental change: Longitudinal approaches to individual adaptation in
childhood. Child Development, 61, 1363-1373.
Steinberg, L., Lamborn, S. D., Darling, N., Mounts, N.S., & Dornbusch, S. M. (1994).
Over-time changes in adjustment and competence among adolescents from authoritative,
authoritarian, indulgent, and neglectful families. Child Development, 65, 754-770.
Tinsley, H. E., & Tinsley, D. J. (1987). Uses of factor analysis in counseling
psychology research. Journal of Counseling Psychology, 34, 414-424.

School Competence and Early Attachment
39
Thompson, R. A. (1999). Early attachment and later development. In: J. Cassidy & P.
R. Shaver, (Eds.) Handbook of attachment: Theory, research and clinical applications (pp.
265-286). New York: The Guilford Press.
Teo, A. Carlson, E, Mathieu, P. Egeland, B., & Sroufe, L. A. (1996). A prospective
longitudinal study of psychological predictors of achievement. Journal of School Psychology,
34, 285-306.
Toth, S. L., & Cicchetti, D. (1996). The impact of relatedness with mother on school
functioning in maltreated children. Journal of School Psychology, 34, 247-266.
Van IJzendoorn, M.H., Sagi, A., & Lambermon, M. W. E., (1992). The multiple
caretaker paradox: Some data from Holland and Israel. In R.C. Pianta (Ed.), Special issue on
Relationships between Children and Non-Parental Adults. New Directions in Child
Development, 57, 5-24.
Van IJzendoorn, M. H. & Sagi, A. (1999). Cross-cultural patterns of attachment:
Universal and contextual dimension. In: J. Cassidy & P. R. Shaver, (Eds.) Handbook of
attachment: Theory, research and clinical applications (pp. 713-734). New York: The
Guilford Press.
Verschueren, K., & Marcoen, A. (1999). Representation of self and socioemotional
competence in kindergartners: Differential and combined effects of attachment to mother and
to father. Child development, 70, 183-201.
Waters, E., & Cummings, M.E. (2000). A secure base from which to explore close
relationships. Child Development, 71, 164-172.
Wechsler, D. (1974). Manual for Wechsler Intelligence Scale for Children-Revised.
New York: The Psychological Corporation.
Weinfield, N.S., Sroufe, A., Egeland, B., & Carlson, E. A. (1999). The nature of
individual differences in infant-caregiver attachment. In J. Cassidy & P. R. Shaver (Eds.),

School Competence and Early Attachment
40
Handbook of Attachment: Theory, research and Clinical Applications (pp 68-88). New York:
The Guilford Press.
Weinfield, N. S., Sroufe, L. A., & Egeland, B. (2000). Attachment from infancy to
early adulthood in a high-risk sample: Continuity, discontinuity, and their correlates. Child
Development, 71, 695-702.
Wentzel, K.R. (1998). Social relationships and motivation in middle school: The role
of parents, teachers, and peers. Journal of Educational Psychology, 90, 202-209.
Wentzel, K. R., Feldman, S. S., & Weinberger, D. A. (1991). Parental child rearing
and academic achievement in boys: The mediational role of social-emotional adjustment.
Journal of Early Adolescence, 11, 321-339.
Zimmermann, P., & Grossmann, K. E. (1997). Attachment and adaptation in
adolescence. In: W. Koops, J.B. Hoeksma, & D.C. van den boom (Eds.). Development of
interaction and attachment: Traditional and non-traditional approaches (pp.281-282).
Amsterdam: North–Holland.

School Competence and Early Attachment
41
Acknowledgements
This study was supported by the United States-Israel Binational Science Foundation
(BSF) grant no. 90-489 to Abraham Sagi, Gary Resnick and Ora Aviezer. Portions of these
data were presented at the biennial meeting of the Society for Research in Child
Development, Albuquerque, NM, 1999. We thank Dr. Michael Nathan, Dr. Arza Avrahami
and Yehudit Binyamin of the Institute for Research on Kibbutz Education at Oranim
Teachers’ College for Help in recruiting the participants and Dr. Miri Scharf for coding the
SAT. We would also like to thank Prof. Rachel Seginer for her helpful comments. Special
thanks are extended to the children, parents and teachers as well as educational coordinator in
the kibbutzim who so generously gave us of their time and cooperation.

School Competence and Early Attachment
42
Table 1: Measures of children’s competence at school derived from teachers’ questionnaire.
Measure
Factor
Loading
% of
Variance
Eigenvalue
Analysis I: Academic Functioning
Scholastic Attitude (alpha reliability = .91)
61
5.37
Task persistence
.90
Motivation
.89
Efficient use of time
.87
Attention span
.68
Attitude to school
.67
Enthusiasm
.66
Scholastic skills (alpha reliability = .83)
13
1.24
Oral verbal skills
.88
Curiosity
.84
Writing skills
.72
Analysis II: Social Emotional Functioning
Social competence (alpha reliability = .81)
43
4.34
Children like him
.84
Has social skills
.84
Able for group work
.69
Behaves emotionally with peers
.67
Emotional Maturity (alpha reliability = .82)
14
1.38
Works independently
.86
Copes with difficulties
.79
Handles criticisms
.75
Behavioral difficulties (alpha reliability = .69)
13
1.32
Needs disciplinary measures
.81
Interruptive behavior in class
.79
Responds non-constructively to failure
.58

School Competence and Early Attachment
43
Table 2: Correlations of infant attachment and SAT scales, sleep, life event clusters, IQ and
perceived competence scales to school adaptation measures.
Outcome measures of school competence
Scholastic
Attitude
Scholastic
Skill
Emotional
Maturity
Social
competence
Behavior
Difficulty
General
Progress
GPA
Att motha
.19
.29*
.23+
.05
-.07
.06
.18
Att fatha
- .12
.16
- .09
.08
.10
.02
.08
LEQ care
- .03
-.02
- .09
.02
.07
.02
.02
LEQspou
- .28*
-.11
- .22+
-.22+
.20
.00
.05
Sleepb
.14
-.09
.28*
.10
-.20
.08
.08
SATscale
.26*
.50**
.22+
.30*
.14
.26*
.32**
IQ
.23+
.48**
.40**
.17
- .09
.46**
.37**
PCAcad
.50**
.57**
.54**
.15
- .28*
.65**
.56**
PCC
.24+
.24+
.24+
.13
- .46**
.14
.15
PCApp
- .17
- .08
-.14
-.12
.15
-.08
- .08
PCAthlet
- .13
.02
.11
-.05
.19+
.06
- .02
PCSelf
.04
.05
-.08
-.03
.02
.04
.03
Gender
.10
- .09
.00
-.10
- .11
- .14
- .12
a1=insecure, 2=secure b1=collective, 2=home; LEQspou – LEQ spouse relation cluster
PCAcad - Perceived Competence Academic, PCC-Perceived Competence Conduct, PCApp -
Perceived Competence Appearance, PCAthlet - Perceived Competence Athletic, Perceived
Competence Self Esteem. All significance levels are 2-tailed.
**p< .01, *p< .05, +p< .10

School Competence and Early Attachment
44
Table 3: Simultaneous Multiple Regression of School Adaptation Measures on Predictor
variables
Academic Oriented Variables
Beta
Total R2
Scholastic Attitude (n=57)
.55**
Infancy Attachment to mother
.199+
LEQ – Spouse Relationship
-.161
SAT
.215+
IQ
-.011
Perceived Academic Competence
.444***
Perceived Behavioral Competence
-.019
Scholastic Skill (n=57)
.65***
Infancy Attachment to mother
.282**
SAT
.437***
IQ
.241**
Perceived Academic Competence
.456***
Perceived Behavioral Competence
.007
General Progress (n=61)
.54***
SAT
.192*
IQ
.256**
Perceived Academic Competence
.569***
Grade Point Average (n=54)
.44***
Infancy Attachment to mother
.137
SAT
.287**
IQ
.168
Perceived Academic Competence
.490***

School Competence and Early Attachment
45
Table 3 continued.
Socioemotional Variables
Beta
Total R2
Emotional Maturity (n=58)
.52***
Infancy Attachment to mother
.258**
LEQ – Spouse Relationship
-.131
Sleeping Arrangement
.368***
SAT
.199*
IQ
.208*
Perceived Academic Competence
.416***
Perceived Behavioral Competence
-.023
Social Competence (n=60)
.19**
Infancy Attachment to mother X Spouse Relationship
.486**
LEQ – Spouse Relationship
-.503**
SAT
.254*
Behavioral Difficulties (n=59)
.27***
LEQ – Spouse Relationship
.147
Sleeping Arrangement
-.170
Perceived Academic Competence
-.102
Perceived Behavioral Competence
-.390**
+p< .10, *p <.05, **p<.01, ***p<.001

School Competence and Early Attachment
46
Table 4: Hierarchical Multiple Regression of School Adaptation Measures on Predictor
variables
Academic Oriented Variables
Total R2
R2 Change
Scholastic Attitude (n=57)
a. IQ
Perceived Academic Competence
Perceived Behavioral Competence
.23***
-
.234***
-
b. SAT
.28***
.046+
c. LEQ – Spouse Relationship
.29***
.014
d. Infancy Attachment to mother
.33***
.036+
Scholastic Skill (n=57)
a. Perceived Academic Competence
IQ
Perceived Behavioral Competence
.42***
-
.315***
.101**
-
b. SAT
.57***
.155***
c. Infancy Attachment to mother
.65***
.076**
General Progress (n=61)
a. Perceived Academic Competence
IQ
.50***
.426***
.076**
b. SAT
.54***
.036*
Grade Point Average (n=54)
a. IQ
Perceived Academic Competence
.31***
-
.310***
b. SAT
.39***
.083**
c. Infancy Attachment to mother
.42***
.024

School Competence and Early Attachment
47
Table 4 continued
Socioemotional Variables
Total R2
R2 Change
Emotional Maturity (n=58)
a. Perceived Academic Competence
IQ
Perceived Behavioral Competence
.32***
.269***
.051*
-
b. SAT
.35***
.027
c. Sleeping Arrangement
LEQ – Spouse Relationship
.45***
.106**
-
d. Infancy Attachment to mother
.50***
.048*
Social Competence (n=60)
a. SAT
.06*
.064*
b. Infancy Attachment to mother X
LEQ-Spouse Relationship
.08+
.013
c. LEQ – Spouse Relationship
.19**
.112**
Behavioral Difficulties (n=59)
a. Perceived Academic Competence
Perceived Behavioral Competence
.21***
-
.210***
b. Sleeping Arrangement
LEQ – Spouse Relationship
+p< .10, *p <.05, **p<.01, ***p<.001