Peer Contagion of Depressogenic Attributional Styles Among Adolescents: A Longitudinal Study

Article · March 2005with167 Reads
DOI: 10.1007/s10802-005-0931-2 · Source: PubMed
Abstract
This study examined longitudinal associations between adolescents' and their friend's depressive symptoms and depressogenic attributional style. Participants included 398 adolescents in grades six through eight at the outset of the study. Adolescents completed peer nominations to identify reciprocated and unreciprocated best friendships as well as measures of depressive symptoms and depressogenic attributional style at an initial time point, and again 11 months later. Results revealed that best friends' reported level of depressive symptoms was prospectively associated with adolescents' own depressive symptoms and with adolescents' depressogenic attributional style. Moderator effects suggested that friends' attributional styles were prospectively associated with adolescents' own attributional styles for those involved in reciprocated friendships. Lastly, findings offered preliminary support for adolescents' Time 2 depressive symptoms as a mediator of the association between friends' depressive symptoms and adolescents' attributional style. Findings have important implications for cognitive and interpersonal models of adolescent depression, as well as the study of peer contagion effects.
Journal of Abnormal Child Psychology, Vol. 33, No. 1, February 2005, pp. 25–37 (
C
2005)
DOI: 10.1007/s10802-005-0931-2
Peer Contagion of Depressogenic Attributional Styles
Among Adolescents: A Longitudinal Study
Elizabeth A. Stevens
1
and Mitchell J. Prinstein
2,3
Received November 24, 2003; revision received May 12, 2004; accepted July 8, 2004
This study examined longitudinal associations between adolescents’ and their friend’s depressive
symptoms and depressogenic attributional style. Participants included 398 adolescents in grades six
through eight at the outset of the study. Adolescents completed peer nominations to identify re-
ciprocated and unreciprocated best friendships as well as measures of depressive symptoms and
depressogenic attributional style at an initial time point, and again 11 months later. Results revealed
that best friends’ reported level of depressive symptoms was prospectively associated with adoles-
cents’ own depressive symptoms and with adolescents’ depressogenic attributional style. Moderator
effects suggested that friends’ attributional styles were prospectively associated with adolescents’ own
attributional styles for those involved in reciprocated friendships. Lastly, findings offered preliminary
support for adolescents’ Time 2 depressive symptoms as a mediator of the association between friends’
depressive symptoms and adolescents’ attributional style. Findings have important implications for
cognitive and interpersonal models of adolescent depression, as well as the study of peer contagion
effects.
KEY WORDS: depression; friendship; attributional style.
Investigations of cognitive theories of depression
among adolescents have now yielded substantial evidence
to suggest that individuals’ attributional style can con-
tribute significantly to the development of depressive
symptoms (Abramson, Metalsky, & Alloy, 1989; Hankin
& Abramson, 2001). Specifically, studies have revealed
that the tendency to attribute negative life events to inter-
nal, global, and stable causes, and positive events to ex-
ternal, specific, and unstable causes (i.e., a depressogenic
attributional style) is significantly associated with adoles-
cents’ depression (see Gladstone & Kaslow, 1995; Joiner
& Wagner, 1995 for reviews) and can predispose individu-
als to future depressive symptoms in the event of stressful
circumstances (e.g., Abela, 2001; Hankin, Abramson, &
Siler, 2001; Panak & Garber, 1992; Prinstein & Aikins,
2004). Thus, recent work has emphasized a need to iden-
1
Yale University, New Haven, Connecticut.
2
Department of Psychology, University of North Carolina, Chapel Hill,
North Carolina.
3
Address all correspondence to Mitchell J. Prinstein, Department of
Psychology, University of North Carolina, Davie Hall, CB #3270,
Chapel Hill, North Carolina 27599-3270; e-mail: mitch.prinstein@
unc.edu.
tify possible antecedents of a depressogenic attributional
style that might be targeted in preventive intervention
efforts. A focus on potential gender differences in the
examination of antecedents of attributional style is espe-
cially important given girls’ greater cognitive vulnerabil-
ity to depression in adolescence (Hankin & Abramson,
2001) and the dramatic increases in depressive symptoms
that emerge among girls during this developmental period
(Hankin et al., 1998).
Few prior studies have examined antecedents of ado-
lescents’ depressogenic attributional style. To date, only
three sets of predictors have been explored. Consistent
with hopelessness theories of depression, one of these an-
tecedents is depressed affect itself. Garber, Keiley, and
Martin (2002) revealed that initial levels of youths’ de-
pressed affect are prospectively associated with increasing
trajectories of youths’ depressogenic attributional style
over a 6-year period (see also Nolen-Hoeksema, Girgus,
& Seligman, 1992). A second area of research has sug-
gested that negative life events or adverse circumstances in
childhood may contribute to the development of a depres-
sogenic attributional style in adolescence. For example,
evidence from both retrospective and prospective studies
25
0091-0627/05/0200-0025/0
C
2005 Springer Science+Business Media, Inc.
26 Stevens and Prinstein
has revealed that a history of physical/emotional/sexual
maltreatment, emotionally manipulative parenting, or ma-
jor life stressors is associated with a depressogenic attribu-
tional style in adolescence and young adulthood
(Crosseld, Alloy, Gibb, & Abramson, 2002; Garber &
Flynn, 2001; Gibb, 2002; Gibb, Abramson, & Alloy, 2004;
Gibb, Alloy, Abramson, & Marx, 2003; Gibb et al., 2001;
Rose & Abramson, 1992).
A third area of research examining antecedents of a
depressogenic attributional style has explored the trans-
mission of depressive cognitions (or symptoms) within
signicant interpersonal relationships. Studies examining
this contagion hypothesis propose that individuals may
be particularly likely to develop a depressogenic attribu-
tional style or depressive symptoms through interactions
with signicant others who exhibit high levels of depres-
sive cognitions or symptoms. For instance, past work has
provided evidence for synchrony in the timing of moth-
ersand childrens depressive episodes (Hammen, Burge,
& Adrian, 1991) as well support for mothers depressive
symptoms as prospective predictors of adolescents de-
pressive symptoms and attributional styles (Downey &
Coyne, 1990; Garber & Flynn, 2001; Hammen & Brennan,
2001; Jaenicke et al., 1987). The adult psychopathology
literature provides evidence for the contagion of depres-
sive symptoms within college roommate dyads (Joiner,
1994) and correlations between the level of depressive
symptoms among spouses (Tambs, 1991). Social psycho-
logical studies using experimental paradigms also have
provided evidence for contagion of sad affect among unfa-
miliar dyads of young adults (e.g., Strack & Coyne, 1983).
The present study focuses specically on the poten-
tial contagion of depressive symptoms and cognitions be-
tween adolescents and their peers. The study of contagion
effects offers an important direction for research on an-
tecedents of a depressogenic attributional style not only
because this provides a possible target for preventive inter-
ventions (i.e., by addressing the reinforcement of depres-
sive cognitions from signicant others), but also to un-
derstand the synergy between cognitive and interpersonal
factors that poses increased risk for depressive symptoms
in adolescence.
Theory and research suggest that signicant increases
in depressive symptoms and cognitive vulnerability that
occur at the transition to adolescence are likely related,
in part, to substantial interpersonal transitions also occur-
ring at this developmental stage, particularly among girls
(Cyranowski, Frank, Young, & Shear, 2000; Hankin &
Abramson, 2001; Rudolph & Hammen, 1999). This crit-
ical developmental period is characterized by signicant
increases in the frequency of youths contact with peers,
increased autonomy from parents, and experimentation
with new roles and experiences in the peer context (Brown,
1990; Rudolph & Hammen, 1999). Given these changes,
adolescents interactions with peers may provide a unique
context for the potential transmission of depressive symp-
toms and depressogenic attributional styles; however, this
hypothesis rarely has been examined. As compared to chil-
dren, adolescents establish friendships that are character-
ized by high levels of intimacy and emotional disclosure,
particularly among girls (Buhrmester & Furman, 1987;
Furman & Buhrmester, 1992). Adolescents also are es-
pecially likely to rely on peers as sources of social sup-
port during times of stress (Harter, Stocker, & Robinson,
1996), and thus opportunities to discuss life stressors and
relevant attributions are numerous (Rose, 2002). Research
also suggests that susceptibility to peer inuence increases
at this developmental period as many adolescents crave
conformity and acceptance within the peer group and use
peer interactions as a basis for reected appraisal of their
own self-worth (Brown, 1990).
Indeed, substantial evidence demonstrates that ado-
lescent adjustment is strongly associated with the social-
psychological adjustment of their closest friends. Effects
of peer contagion have been revealed in the study of ado-
lescents delinquency, illegal behavior, aggression, and
a host of health risk behaviors, including substance use,
sexual risk behavior, general risk behavior, and suicidal-
ity (e.g., Kandel, 1996; Prinstein, Boergers, & Spirito,
2001; Prinstein, Meade, & Cohen, 2003; Urberg,
Degirmencioglu, & Pilgrim, 1997; Vitaro, Tremblay, Kerr,
Pagani, & Bukowski, 1997; Wills & Cleary, 1999). Find-
ings have revealed that adolescent peer contagion is not
only due to selection of friends with similar social-psych-
ological characteristics (i.e., selection effects), but also to
socialization effects (Kandel, 1978). Although the mecha-
nisms responsible for socialization effects remain largely
unknown, results have demonstrated that characteristics
of adolescents friends are signicantly and prospectively
associated with increases in adolescentsown symptoms
and behaviors (e.g., Keenan, Loeber, Zhang, Stouthamer-
Loeber, & Van Kammen, 1995).
Some evidence is available to suggest that peer-
contagion effects may apply to the homophily of inter-
nalizing symptoms (e.g., depression) specically. Several
studies have revealed that among adults, depressed indi-
viduals are particularly likely to select other depressed
individuals as friends and respond more favorably to so-
cial interactions that involve a similarly depressed partner
(Rosenblatt & Greenberg, 1988, 1991). Evidence for a
socialization effect within youth samples has been pro-
vided in a recent study examining peer cliques (Hogue
& Steinberg, 1995). Findings revealed that adolescents
own internalizing distresswas predicted longitudinally
Peer Contagion 27
by the average level of internalizing distress within their
friendship group.
The current study was designed to extend ndings re-
garding antecedents of a depressogenic attributional style,
as well as research regarding the effects of peer contagion,
by examining associations between adolescents and their
best friendsattributional style and depressive symptoms.
The focus on adolescents best friends allowed for a study
of depressogenic attributional style antecedents that relied
on external reporters (i.e., best friends). This study also
utilized a prospective, longitudinal design among adoles-
cents who were at the critical developmental stage during
which cognitive vulnerabilities to depressive symptoms
may emerge.
To replicate and extend ndings reported by Hogue
and Steinberg (1995), an initial hypothesis suggested that
best friends depressive symptoms and attributional style
would be prospectively associated with adolescentsown
depressive symptoms. It also was predicted that best
friends depressive symptoms and attributional style
would be prospectively associated with adolescentsown
depressogenic attributional style. To provide a stringent
test of these hypotheses, the incremental contributions of
friends depressive symptoms and attributions were ex-
amined after controlling for adolescents initial levels of
depressive symptoms and attributional styles as predictors.
The potential effects of friends depressive symptoms and
attributional styles were examined simultaneously to de-
termine whether each contributed uniquely to adolescents
own functioning. As a marker of sad affect or low levels
of positive reinforcement within friendships, friendsde-
pressive symptoms may be most relevant as an agent of
contagion between friends. As an indicator of friends po-
tential reinforcement of maladaptive cognitions, friends
attributional styles may be a more relevant predictor of
adolescentsown functioning.
Two potential moderators of these predicted conta-
gion effects were examined. Consistent with prior nd-
ings regarding greater similarities between mutual (i.e.,
reciprocated) as compared to unilateral (i.e., unrecipro-
cated) friends (e.g., Haselager, Hartup, van Lieshout, &
Riksen-Walraven, 1998; Kandel, 1978; Lea, 1979), it was
hypothesized that friendship reciprocity would be a signif-
icant moderator of the association between adolescents
and their friendsdepressive symptoms. Stronger associ-
ations among reciprocal friends might be due to higher
levels of friendship quality and increased opportunities
for interaction and disclosure than among unreciprocated
friends (Parker & Asher, 1993). Gender was examined
as a second moderator. Research and theory indicate that
girls are not only at greater risk for developing depressive
symptoms and cognitive vulnerabilities during the transi-
tion to adolescence, but also may exhibit unique behaviors
that are especially likely to contribute to the contagion
of depressive symptoms and depressogenic attributions
within close friendships (e.g., co-rumination; Rose, 2002;
see below). Moreover, girls are more likely to develop re-
ciprocated friendships than boys (Parker & Asher, 1993).
Thus, it was anticipated that gender and friendship reci-
procity may be competing moderators that explain similar
variance in peer contagion.
In addition to an examination of the presence of po-
tential contagion effects, this dataset also offered an op-
portunity to examine mediator models, helping to eluci-
date the direction of effects involved in the contagion of
depressive symptoms and attributional styles. An initial
model suggested that best friends depressive symptoms
and/or attributional styles might exert a direct inuence on
adolescents own attributional style, with indirect impli-
cations for adolescentsown depressive symptoms (i.e., a
cognitive inuence model). In other words, it was antici-
pated that the association between best friends depressive
symptoms (or attributional style) and adolescents own
depressive symptoms would be mediated by adolescents
own depressogenic attributional style.
Recent research regarding unique depression-related
friendship behaviors, particularly among adolescent girls,
offers compelling evidence for the effects of best friends
depressive symptoms and/or cognitions on adolescents
attributional styles directly. Rose (2002) indicates that
high levels of friendship quality are positively associated
with repeated discussions of interpersonal ambiguities and
excessive self-disclosure between friends (i.e., referred to
as co-rumination). Co-rumination is more commonly
reported among girls than boys, and is a signicant me-
diator of the gender disparity in the prevalence of adoles-
cent internalizing symptoms (Rose, 2002). Through co-
rumination, best friends have an opportunity to explicitly
discuss attributions of interpersonal events; thus, direct
effects of friends depressive symptoms or attributional
styles on adolescents own depressive attributional style
may be especially likely, and may explain the transmis-
sion between the level of depressive symptoms among
adolescents and their best friends.
An alternate mediator model suggests an indirect
contagion effect of depressive attributions via depressive
symptoms. Specically, it may be that best friends
depressive symptoms and/or attributional styles are asso-
ciated longitudinally with adolescents depressive symp-
toms, and adolescents symptoms serve as a proximal
antecedent to their own subsequent depressive cognitions.
Several possible theoretical models may be rele-
vant for understanding this potential symptom inuence
model.Research in social psychology demonstrates that
28 Stevens and Prinstein
individuals depressed affect increases through interac-
tions with others who exhibit sad affect, as would be pre-
dicted by social learning theories (Strack & Coyne, 1983).
Symptoms of depression also might be potentially trans-
mitted between friends through mutual interactions that
lack sufcient positive reinforcement or adaptive social
support (Lewinsohn, 1974). Although these symptom and
cognitive inuence models might best be explored over
three or more time points, prospective analyses of depres-
sive symptoms and cognitions as both predictors and/or
mediators in this study offers a preliminary opportunity
to examine directions of effect between two inter-related
constructs and explore mechanisms that might explain
peer contagion.
METHOD
Participants
Participants included 398 children and adolescents
(48% female) in grades six (36%), seven (27%), and eight
(37%) at the outset of the study. All participants were
between the ages of 11 and 14 years old, M = 12.70;
SD = .95. The ethnic composition of the sample included
87% White/Caucasian, 2% African-American, 4%
Asian-American, 2% Latino-American, and 6% of par-
ticipants from multi-ethnic backgrounds. At Time 1, par-
ticipants were enrolled in one middle school within a city
of fairly homogeneous middle-class socioeconomic sta-
tus. According to school records, 11% of children were
eligible for free/reduced-fare lunch.
At the outset of the study all sixth, seventh, and
eighth grade students were recruited for participation (i.e.,
approximately 300 students per grade). Consent forms
were returned by 92% of families (n = 846); of these,
80% of parents gave consent for their childs participa-
tion (n = 677; 74% of the total population), and all but 52
of these children provided information at Time 1. Miss-
ing data were due to absenteeism during one of the days
of testing (n = 35), incomplete responses (n = 13) and
denial of assent to participate in the study (n = 4).
Because a main focus of this study was to exam-
ine associations between adolescents and their friends
depressive cognitions and symptoms, all analyses were
restricted to adolescents who selected friends for whom
data were also available. Of the 625 youth who partici-
pated at Time 1 of this study, 462 selected a best friend
who was also a study participant. Chi-square and t-test
analyses revealed no signicant differences on any study
variables between adolescents who did vs. did not select a
best friend who was also a study participant. Of these 462
adolescents with available data, 20 selected a best friend
who also was selected by at least one other participant,
thus 12 of these were deleted at random to ensure that
friends data were included only once in analyses.
Of these 450 adolescents, a total of 398 (88%) were
available for testing 11 months later (i.e., Time 2), when
students were in grades seven through nine. Ninth grade
students had entered the local high school; however, no
interactions by grade were revealed in analyses of main
study hypotheses below. Attrition was due to participants
moving away from the area (n = 36), absenteeism (n =
7), incomplete data (n = 9), and ve participants who re-
fused to continue participation. Attrition analyses revealed
no signicant differences on any of the measures included
in this study between the 398 adolescents in the nal sam-
ple as compared to the remaining adolescents from the
initial sample.
Measures
Depressive Symptoms
The Childrens Depression Inventory (CDI; Kovacs,
1992) is a 27-item questionnaire that assesses cognitive
and behavioral depressive symptoms. A three-item re-
sponse format is used, scored 02, in which children en-
dorse statements that best describe their level of depressive
symptoms in the previous 2 weeks. A summed score was
computed with a higher score indicating a higher level of
depressive symptoms. Good psychometric properties have
been reported for the CDI as a reliable and valid index of
depressive symptoms (Saylor, Finch, Spirito, & Bennett,
1984). In the present sample internal consistency was high
(Time 1 and Time 2 αs = .88).
Attributional Style
The 24-item version of the childrens attributional
style questionnaire (CASQ) was used as a measure of
depressogenic attributional style at Time 1 and Time 2
(Thompson, Kaslow, Weiss, & Nolen-Hoeksema, 1998).
This measure assesses three attributional dimensions
(internalexternal, stableunstable, and globalspecic)
that have been associated with depression (Abramson
et al., 1989). The CASQ lists 12 positive (e.g., athletic suc-
cess, praise) and 12 negative (e.g., academic failure, inter-
personal rejection) events; for each, adolescents are asked
to select one of two possible causes. For both positive
and negative events, a score of 1 was coded for each in-
ternal, stable, or global attribution, and a score of 0 was
coded for each external, unstable, or specic attribution.
Consistent with prior research (Panak & Garber, 1992;
Seligman et al., 1984), a sum of attributions for negative
Peer Contagion 29
events was subtracted from a sum of attributions for posi-
tive events to derive a overall summary score, with lower
total CASQ scores indicating higher levels of depresso-
genic attributional style. Results from the current sample
suggested adequate internal consistency for this measure
(i.e., Time 1 = .66; Time 2 = .73), consistent with pre-
vious investigations utilizing the CASQ (Garber et al.,
2002; Gladstone & Kaslow, 1995; Robins & Hinkley,
1989).
Friendship Selection
A peer nomination procedure was used to identify
adolescents closest friends at Time 1 (Parker & Asher,
1993). Adolescents were asked to select an unlimited num-
ber of their closest friends from a roster of grademates
alphabetized by rst name (or nickname, e.g., Bill/
William, as indicated in focus groups), and from this list to
select a very best friendas well as two additional best
friends. Adolescents very best friends scores on the CDI
and the CASQ were used as a measure of friends de-
pressive symptoms and depressogenic attributional style,
respectively.
The reciprocity of adolescents very best friend se-
lection was determined by examining mutual friendship
nominations. Adolescents very best friendship nomina-
tion was coded as a reciprocal best friendship if the peer
selected as an adolescents very best friend selected the
adolescent as one of their best friends (i.e., among the top
3 closest friends) (Parker & Asher, 1993). Of the 450 ado-
lescents at Time 1 for whom reciprocity data were avail-
able, 313 adolescents (69%) very best friendship choices
were reciprocated best friendships at Time 1. Of the
remaining 137 adolescents, 93 adolescents (21% of the
sample) selected a very best friend who selected the ado-
lescent as a close friend, but not a best friend (i.e., not
among the top 3 closest friends), and 44 adolescents (10%
of the sample) selected a very best friend who did not se-
lect the adolescent as a friend at all. These relationships
were coded as reciprocal friendships and unrecipro-
cated friendships,respectively.
4
4
In the friendship literature, reciprocal friendship coding is often used
to determine the number of adolescentsreciprocated friendships. For
these purposes, adolescentssecond and third choice friend selections
also might be referred to as reciprocal best friends if these peers re-
ciprocated adolescentsnomination (e.g., Parker & Asher, 1993). This
study focused specically on associations between adolescents and
their very best friends (i.e., only the number one closest friends) psy-
chological functioning, consistent with prior literature examining peer
contagion effects. Thus, the use of the term reciprocal friendship in
this study is not meant to capture the number of adolescents reciprocal
friendships, but rather to serve as an indicator of whether the peer that
adolescents selected as a very best friend (i.e., listed as the number one
Table I. Descriptive Statistics for Main Study Variables, Means and
Standard Deviations (n = 398)
Mean SD Observed range
Time 1 Variables
Depressive symptoms 7.47 6.49 033.00
Attributional style 5.99 3.19 6.0012.00
Time 2 Variables
Depressive symptoms 6.47 6.08 032.00
Attributional style 6.40 3.67 5.0012.00
p <.05.
Chi-square analyses revealed that boys were over-
represented within the group of adolescents whose very
best friend nominations were not reciprocated at all (i.e.,
unreciprocated friendships), χ
2
(2) = 12.82, p <.001.
No other signicant differences emerged on any mea-
sures in this study between adolescents in reciprocal best
friendships, reciprocal friendships, and unreciprocated
friendships.
RESULTS
Preliminary Analyses
Means, standard deviations, and observed ranges for
all primary variables are presented in Table I. As would
be expected in this community sample, the prevalence of
clinically signicant levels of depressive symptoms (i.e.,
t-scores above 63) was approximately 5% at both time
points (Time 1 = 6.4%; Time 2 = 4.0%). Preliminary
analyses revealed no gender or grade main effects in par-
ticipants reports of depressive symptoms or depresso-
genic attributional styles at Time 1 or Time 2. Two sta-
tistically signicant grade × gender interaction effects
were revealed; however, neither accounted for a mean-
ingful effect, η
2
<.03. Simple effects analyses revealed a
signicant grade effect for boys only, F(2, 201) = 5.17,
p <.05, indicating that students in the seventh grade at
Time 2 reported lower levels of Time 2 depressive symp-
toms, M = .17; SD = .19, than eighth grade students,
M = .29; SD = .26; no signicant differences emerged
for students in the ninth grade, M = .24; SD = .19. For
girls only, a signicant grade effect revealed that students
in the seventh grade at Time 2 reported lower Time 2
CASQ scores, M = 5.92; SD = 4.10, than eighth grade
students, M = 7.74; SD = 2.85, F(2, 191) = 4.12, p <
.05; no signicant differences emerged for students in the
ninth grade, M = 6.60; SD = 3.49.
best friend) also reported that the adolescent was a best friend, a close
friend, or not a friend at all.
30 Stevens and Prinstein
Table II. Correlations among Adolescents and their Friends Depressive Symptoms and Attributional Style
Time 1 Time 2
Adolescents Friends Friends Adolescents Adolescents
CASQ CDI CASQ CDI CASQ
Time 1
CDI .56
∗∗
.16
∗∗
.19
∗∗
.66
∗∗
.50
∗∗
CASQ .20
∗∗
.18
∗∗
.44
∗∗
.62
∗∗
Friends CDI .59
∗∗
.32
∗∗
.23
∗∗
Friends CASQ .20
∗∗
.19
∗∗
Time 2
CDI .63
∗∗
Note.
p <.001.
∗∗
p <.0001. CDI = Depressive symptoms; CASQ = Attributional style.
Pearson correlations were conducted to examine as-
sociations among primary variables (see Table II). Re-
sults revealed moderate stability of adolescents depres-
sive symptoms and depressogenic attributional style from
Time 1 to Time 2. Results also revealed signicant associ-
ations between adolescents and their friendsdepressive
symptoms and depressogenic attributional style, both con-
currently and longitudinally.
Longitudinal Prediction of AdolescentsDepressive
Symptoms from Friends Depressive Symptoms
and Attributional Style
It was predicted that best friends depressive symp-
toms and attributional style would be associated longitu-
dinally with adolescents own depressive symptoms. A hi-
erarchical multiple regression analysis using adolescents
depressive symptoms at Time 2 as a dependent variable
was conducted to examine this hypothesis. Adolescents
depressive symptoms and attributional style at Time 1
were entered on an initial step. The entry of adolescents
attributional style as a predictor allowed for an examina-
tion of the incremental predictive value of friends depres-
sive symptoms and attributional style above and beyond
the effects of adolescents own attributional style. This
approach also helped to minimize the possibility that ado-
lescents friends cognitions and symptoms were merely
serving as a proxy for adolescents own attributional style
as a predictor. Dummy-coded variables were entered on a
second step to account for the main effects of gender and
the reciprocity of adolescents best friend nominations.
For the three levels of friend reciprocity (i.e., recipro-
cal best friendship, reciprocal friendship, unreciprocated
friendship), two dummy-coded variables were computed
(Cohen & Cohen, 1983). The rst of these variables, reci-
procity 1 (i.e., reciprocated best friendships vs. others),
was coded to reveal the effects of reciprocated best friend-
ships as compared to the other two groups combined, and
the second variable, reciprocity 2 (unreciprocated friend-
ships vs. others) was coded to reveal the effects of unrecip-
rocated friendships as compared to the other two groups
combined.
To examine the main study hypothesis, adolescents
friends depressive symptoms and attributional style were
entered on a third step. Product terms were computed be-
tween each predictor (friends attributional style and de-
pressive symptoms) with each of the three dummy-coded
variables (i.e., gender, reciprocity 1, reciprocity 2). Be-
cause of signicant associations between each of these
predictors and between each of the potential moderators,
there was substantial multicollinearity between the result-
ing six product terms. Moderators were therefore exam-
ined separately in two parallel analyses. In the rst anal-
ysis, product terms examining gender and reciprocity as
moderators of best friends attributional style were entered
on Step 4. The analysis was then reconducted using prod-
uct terms examining gender and reciprocity as moderators
of friends depressive symptoms on Step 4; see Table III).
Results revealed that higher levels of friends depres-
sive symptoms at Time 1 were associated longitudinally
with higher levels of adolescents depressive symptoms at
Time 2, after controlling for initial levels of adolescents
depression and attributional style (see β at step). The effect
of friends depressive symptoms was not qualied by ei-
ther gender or friendship reciprocity status; no interaction
terms were statistically signicant (see Alternate Step 4,
Table III). No signicant main or interaction effects were
revealed for friends attributional style as a prospective
predictor of adolescentsown depressive symptoms.
Longitudinal Prediction of Adolescents Attributional
Style from Friends Depressive Symptoms
and Attributional Style
A second hypothesis pertained to the examina-
tion of adolescents friends depressive symptoms and
Peer Contagion 31
Table III. Longitudinal Prediction of Adolescents Time 2 Depressive Symptoms from Friends
Time 1 Depressive Symptoms and Attributional Style
Predictors R
2
β at step Final β
Step 1 .44
Time 1 depressive symptoms .61
.62
Time 1 attributional style .09
.06
Step 2 .00
Gender (Female) .06 .23
∗∗
Reciprocity 1 (reciprocated best friendships vs. others) .00 .11
Reciprocity 2 (unreciprocated friendships vs. others) .00 .13
Step 3 .04
Time 1 friends depressive symptoms .24
.24
Time 1 friends attributional style .07 .27
Step 4 .01
Friends attributional style × gender .36
∗∗
Friends attributional style × reciprocity 1 .01
Friends attributional style × reciprocity 2 .07
Alternate step 4 .00
Friends depressive symptoms × gender .22
Friends depressive symptoms × reciprocity 1 .01
Friends depressive symptoms × reciprocity 2 .00
Note.
p <.001.
∗∗
p <.05.
p = .07.
attributional style as longitudinal predictors of adoles-
cents attributional style. A hierarchical multiple regres-
sion using Time 2 attributional style scores as a dependent
variable was conducted. Adolescents levels of depres-
sogenic attributions and depressive symptoms at Time 1
were entered on an initial step. As mentioned earlier, ado-
lescents depressive symptoms were entered as a predictor
on an initial step to examine the incremental contributions
of friends depressive symptoms and attributional style as
predictors and to ensure that friends depressive symptoms
and cognitions were not serving as a proxy for adolescents
own depression. Dummy-coded variables for gender and
adolescents friendship reciprocity were entered on a sec-
ond step. The main effects of friends depressive symp-
toms and cognitions were entered on Step 3. As above,
product terms were entered on a last step in two parallel
analyses to examine gender and friendship reciprocity as
potential moderators of friends depressive symptoms and
attributional style, respectively (see Table IV).
A main effect was revealed for friends Time 1 de-
pressive symptoms as a predictor of adolescentsattribu-
tional style at Time 2, after controlling for adolescents
initial levels of attributional style and the effects of ado-
lescents own depressive symptoms as a predictor (see β at
step).
5
However, this signicant effect was qualied by a
5
Because both adolescents and their friends were drawn from the same
dataset, there was some potential for data redundancy (i.e., some par-
ticipants were included as both a target adolescent and a best friend
in analyses). To ensure that the magnitude of associations were not
signicant interaction term, suggesting that gender moder-
ated this association. The nature of this moderating effect
was subsequently examined using Holmbecks (2002) re-
cent guidelines for post hoc probing. This included: (a)
recomputation of reducedregression models including
only signicant predictors to eliminate potential errors in
parameter estimation or errors in partialling of unique ef-
fects due to multicollinearity; (b) computation of slope
estimates using centered variables (thus, further reducing
multicollinearity); and (c) examining the statistical signif-
icance of these slopes at different levels of the moderator
variable (e.g., for girls and for boys). A signicant slope
was revealed for girls, b =−2.80; β =−.19, p <.01,
suggesting that higher levels of friends depressive symp-
toms were associated with higher levels of depressogenic
attributional styles (i.e., lower CASQ scores). In contrast,
inated due to data redundancy, all analyses were reconducted using
a reduced dataset in which one participants data from each dyad was
omitted at random and all participants therefore were included as only
a target adolescent or a best friend, but not both. Results from these
analyses revealed an identical pattern of ndings as reported in the text
with one exception. In contrast to ndings reported in Table IV, the
main effect of friends depressive symptoms as a longitudinal predic-
tor of adolescentsown depressogenic attributional style was no longer
signicant. However, the gender × friends depressive symptoms inter-
action effect remained signicant in this analysis, R
2
= .01, p <.01.
As reported for the full sample, the longitudinal association between
best friendsdepressive symptoms at Time 1 and adolescentsown de-
pressogenic attributional style at Time 2 was statistically signicant
for girls, b = 2.10, β =−.15, p <.05, but not for boys, b = 1.20;
β = .07, ns.
32 Stevens and Prinstein
Table IV. Longitudinal Prediction of Adolescents Time 2 Attributional Style from Their Friends
Time 1 Depressive Symptoms and Attributional Style
Predictors R
2
β at step Final β
Step 1 .42
Time 1 attributional style .49
.46
Time 1 depressive symptoms .22
.23
Step 2 .01
Gender (female) .07 .16
∗∗
Reciprocity 1 (reciprocated best friendships vs. others) .01 .01
Reciprocity 2 (unreciprocated friendships vs. others) .04 .05
Step 3 .01
∗∗
Time 1 friends depressive symptoms .10
.01
Time 1 friends attributional style .00 .12
Step 4 .01
∗∗
Friends attributional style × gender .20
Friends attributional style × reciprocity 1 .13
Friends attributional style × reciprocity 2 .25
∗∗
Alternate step 4 .01
∗∗
Friends depressive symptoms × gender .26
∗∗
Friends depressive symptoms × reciprocity 1 .01
Friends depressive symptoms × reciprocity 2 .11
Note.
p <.05.
∗∗
p <.001.
the slope for boys was not statistically signicantly differ-
ent from zero, b =−.42; β =−.03, ns.
No main effect was revealed for friends attributional
style as a predictor of adolescents own attributional style.
However, a signicant product term revealed that friend-
ship reciprocity (i.e., reciprocity 2) signicantly moder-
ated the longitudinal association between friends and
adolescents attributional style. Post hoc probing revealed
that among adolescents with reciprocated best friendships
and reciprocated friendships (combined), b = .11;
β = .09, p <.05, friends attributional style was asso-
ciated longitudinally with adolescents own attributional
style. No signicant slope emerged for adolescents in un-
reciprocated friendships, b =−.33; β =−.28, ns.
Examination of a Cognitive Inuence Model
Two mediator models were proposed. The rst model
suggested that the association between friends Time 1 de-
pressive symptoms or attributional style and adolescents
own Time 2 depressive symptoms would be mediated by
adolescentsown Time 2 attributional style. This cogni-
tive inuence model therefore posited that friends func-
tioning may exert a direct inuence on adolescents cogni-
tions, which subsequently are associated with adolescents
depressive symptoms. Analyses reported above support
the preliminary criteria required to examine this mediator
model (Baron & Kenny, 1986; Holmbeck, 2002). Cor-
relations revealed a signicant association between ado-
lescentsTime 2 attributional style and depressive symp-
toms. Analyses above also revealed that friends depres-
sive symptoms were associated prospectively with ado-
lescents own depressive symptoms, for boys and girls,
and with attribution style, for girls only. This mediator
hypothesis was therefore examined for girls (n = 194).
Analyses also revealed that friends attributional style
was signicantly associated with adolescents own Time 2
attributional style among those in reciprocal friendships.
However, no signicant association was revealed with
adolescentsown depressive symptoms. Thus, criteria for
mediation were not met.
A hierarchical multiple regression was conducted to
examine adolescents Time 2 attributional style as a poten-
tial mediator of the association between friends Time 1
depressive symptoms and adolescents Time 2 depressive
symptoms. Because an ideal examination of this mediator
hypothesis would involve data at three time points, re-
sults should be interpreted cautiously. Using adolescents
Time 2 depressive symptoms as a dependent variable,
and initial levels of depressive symptoms and attributional
style entered on an initial step, friends depressive symp-
toms were entered on a second step. Adolescents Time 2
attributional style was next examined on a third step, and
the change in the value of the regression weight for friends
depressive symptoms at Steps 2 and 3 were compared to
test for mediation.
As can be seen in Table V, friends depressive symp-
toms were signicantly associated with adolescents de-
pressive symptoms both before (see β at step) and after
Peer Contagion 33
Table V. Cognitive Inuence Mediator Model: Longitudinal Prediction
of Adolescents Time 2 Depressive Symptoms from Friends Time 1
Depressive Symptoms via Adolescents Time 2 Attributional Style for
Girls in Reciprocal Friendships (n = 194)
Predictors R
2
β at step Final β
Step 1 .51
∗∗
Time 1 depressive symptoms .67
∗∗
.57
∗∗
Time 1 attributional style .07 .16
Step 2 .06
∗∗
Time 1 friends depressive symptoms .25
∗∗
.18
∗∗
Step 3 .06
∗∗
Time 2 attributional style .35
∗∗
Note.
p <.001.
∗∗
p <.0001.
(see nal β) adolescents Time 2 attributional style was
entered into the model. Moreover, the reduction in the
beta weight for friends depressive symptoms between
Steps 2 and 3 was not statistically signicant, z = 1.88,
ns (Holmbeck, 2002). Thus, results did not support the
cognitive inuence model; friends depressive symp-
toms were prospectively associated with adolescents own
depressive symptoms, but not via effects on adolescents
attributional style.
Examination of a Symptom Inuence Mediator Model
The second mediator model proposed that peer con-
tagion would occur at the symptom level. Specically,
this hypothesis suggested that friends depressive symp-
toms would be directly associated with adolescentsown
depressive symptoms, which would subsequently be as-
sociated with increases in adolescentsattributional style.
This model reversed the mediator and criterion variables
in the analysis above, and therefore all preliminary criteria
for a test of mediation were still met.
A hierarchical multiple regression was conducted
with adolescents Time 2 attributional style scores en-
tered as a dependent variable. After controlling for prior
levels of adolescents attributional style and the effects
of adolescents Time 1 depressive symptoms as a pre-
dictor, friends depressive symptoms were entered on a
third step followed by adolescents own Time 2 depres-
sive symptoms on a fourth step. The change in the value
of the regression weight for friends depressive symptoms
at Steps 3 and 4 were compared to test for mediation (see
Table VI).
Higher levels of friends depressive symptoms were
signicantly associated with higher levels of a depres-
sogenic attributional style (i.e., lower CASQ scores) be-
fore, but not after adolescents depressive symptoms were
entered into the model. Moreover, the reduction in beta
Table VI. Symptom Inuence Mediator Model: Longitudinal Predic-
tion of Adolescents Time 2 Attributional Style from Friends Time 1
Depressive Symptoms via Adolescents Time 2 Depressive Symptoms
for Girls in Reciprocal Friendships (n = 194)
Predictors R
2
β at step Final β
Step 1 .51
∗∗
Time 1 attributional style .54
∗∗
.49
∗∗
Time 1 depressive symptoms .25
∗∗
.01
Step 2 .03
Time 1 friends depressive symptoms .19
∗∗
.10
Step 3 .06
∗∗
Time 2 depressive symptoms .37
∗∗
Note.
p <.001.
∗∗
p <.0001.
weight for friends depressive symptoms was statistically
signicant, z = 3.80, p <.01. Analysis of direct and indi-
rect effects indicated that 74% of the effect of friends de-
pressive symptoms on adolescents attributional style was
accounted for by the mediating effects of adolescents de-
pressive symptoms (Holmbeck, 2002). Thus, consistent
with a symptom inuence model, results suggested that
friends depressive symptoms were proximally associated
with adolescentsdepressive symptoms, which were then
associated with adolescentsattributional style.
DISCUSSION
Recent ndings regarding substantial increases in the
prevalence of depressive symptoms during adolescence as
well as the important predictive role of depressive cog-
nitions in the onset and maintenance of adolescent de-
pression have led to an increased interest in possible an-
tecedents of adolescents depressogenic attributional style
(Garber & Flynn, 2001). To date, few investigations in this
area have utilized a prospective longitudinal design, ex-
amined predictors that may explain unique variance in
attributional style beyond the effects of adolescentsown
depressive symptoms, or explored aspects of adolescents
peer relationships that may be relevant to the development
of depressive cognitions.
On the basis of theory and research indicating that
peer relationships may be particularly relevant for under-
standing adolescents vulnerability to depression, espe-
cially among girls (Rudolph, 2002), this study revealed
that the level of depressive symptoms reported by ado-
lescents friends was prospectively and uniquely asso-
ciated with adolescents own depressive symptoms and
girls depressogenic cognitions. Moreover, results indi-
cated that among adolescents in reciprocal friendships,
friends reports of their attributional style were associ-
ated longitudinally with adolescents own attributional
34 Stevens and Prinstein
style. Findings contribute to literatures on cognitive and
interpersonal theories of depression as well as peer-
contagion effects, and offer important directions for
preventive interventions.
Cognitive theories of depression suggest that ado-
lescents tendencies to interpret negative life events as
due to internal, global, and stable causes are directly as-
sociated with the experience of depressed affect, and an
increased risk for depressive disorders (Abramson et al.,
1989). However, given the negative emotional con-
sequences of this attributional style, it is unclear why
individuals continue to rely on depressogenic interpreta-
tions of events. From a behavioral perspective, it might be
predicted that this attributional style would extinguish in
the absence of positive reinforcement. Yet, research has
clearly demonstrated that individuals depressive cogni-
tions are indeed stable over time and applied with mod-
erate stability across domains of life events (Hankin &
Abramson, 2001), suggesting that this cognitive style is
likely maintained by numerous factors. Some of these fac-
tors may pertain to the reciprocal associations between
individuals cognitions and their social experiences. De-
pressed cognitions (e.g., I am worthless) alter individu-
als interpersonal behavior (e.g., social withdrawal), which
reciprocally reinforces the negative cognitions (e.g., Iam
alone, I must be worthless) (e.g., Sacco, 1999). Depres-
sive symptoms and cognitions may also be maintained and
reinforced through interaction with other depressed indi-
viduals (Downey & Coyne, 1990; Hammen & Brennan,
2001). Negative interpretations of life events may be ex-
plicitly discussed and corroborated by others (Rose, 2002),
and depression-motivated socialbehavioral decisions
may be learned by observing others make similar
decisions.
Consistent with these latter hypotheses, ndings from
this study supported a peer-contagion effect for adoles-
cent depressive symptoms and cognitions. The ndings
may be best interpreted to suggest that exposure to a
close friend who exhibits depressive symptoms is a factor
that serves to reinforce or maintain adolescents tenden-
cies towards depression. It is not possible to determine
from these data whether adolescents friends depressive
symptoms are causally associated with adolescents own
depressive symptoms, or predict the development of at-
tributional styles beyond the effects of other potential
predictors (e.g., traumatic events). Nor do these data ad-
dress whether adolescents might initially select friends
who posses similar attributional styles or depressive symp-
toms (i.e., selection effects). However, ndings do support
a socialization (i.e., inuence) effect between friends over
time that might be especially important to consider as a
factor that affects trajectories of depression during this
critical developmental interval and may interfere
with treatment attempts to reduce adolescents negative
cognitions.
Considering adolescents increased orientation to-
wards peer relationships, and heightened susceptibility to
peer inuence during this developmental period, the potent
prospective effects of adolescentsfriends psychological
characteristics on adolescents own functioning are not
entirely surprising. It is especially noteworthy that peer-
contagion effects were revealed in the study of internaliz-
ing symptoms, however. Prior research examining adoles-
cent peer contagion has almost exclusively explored the
transmission of externalizing and health risk behaviors
(e.g., Kandel, 1996). Findings from this study may of-
fer some insight into mechanisms that might explain pro-
cesses of peer contagion more broadly. A straightforward
interpretation of these results would suggest that adoles-
cents depressive symptoms and cognitions are directly in-
uenced by the presence of similar affective states among
best friends. This interpretation is based on the assumption
that adolescents actively discuss their interpretations of
life events and display their emotional states among their
peers. Several ndings support this interpretation. Ado-
lescents indeed develop friendships that are characterized
by higher levels of emotional disclosure, intimacy, and
support than in childhood (Buhrmester & Furman, 1987).
Thus, opportunities to discuss negative affective states and
cognitions are available. These relationship characteristics
are particularly prominent in reciprocated best friendships
and among adolescent girls in particular. Accordingly, re-
sults from this study revealed that contagion effects be-
tween adolescents friends depressive symptoms and ado-
lescentssubsequent attributional style were signicantly
stronger among girls and potential contagion of attribu-
tional styles was evident among adolescents in recipro-
cated friendships.
An alternate interpretation of these results suggests
that afliation with a friend who exhibits depressive symp-
toms increases the risk of other aversive experiences that
can exacerbate adolescents own depressive symptoms or
cognitions. For example, depressed adolescents who afl-
iate with depressed friends may be less likely than other
depressed adolescents to be included in social activities.
Friendships among depressed adolescents may include
lower levels of positive reinforcement and social support
than other friendships. Afliation with depressed friends
also might increase adolescents risk of engagement in
maladaptive behaviors that often occur in collusion with
others (e.g., substance use; school truancy). These poten-
tially mediating variables may explain an indirect effect
of peer afliation that leads to exacerbations of depressive
symptoms.
Peer Contagion 35
Still, a third possibility suggests that friends de-
pressive symptoms serve as merely a marker for other
aspects of adolescents socialpsychological functioning
that may exert a more causal inuence on the develop-
ment of adolescent depression. Future work examining
friends depressive symptoms in conjunction with other
possible antecedents of attributional style (e.g., negative
life events, poor parenting, maternal depression) will be
needed to determine the unique effects of friends depres-
sion on adolescents developing depressive symptoms and
cognitions.
Although this study did not examine these specic
mechanisms, analysis of mediational models offered some
preliminary insights to help determine potential pathways
of transmission between friends and adolescents depres-
sive symptoms and cognitions. Specically, after revealing
an association between friends depressive symptoms and
adolescents own depressive symptoms and cognitions,
it was of interest to explore whether friends symptoms
might have more direct implications for the development
of similar symptoms among adolescents (with indirect
consequences on adolescents attributional style) (i.e.,
symptom inuence model) or direct consequences for the
development of a negative attributional style (with indi-
rect consequences on adolescentsdepressive symptoms)
(i.e., cognitive inuence model). Preliminary results ex-
amining data from two time points among girls offered
compelling support for the symptom inuence model and
no signicant support for the cognitive inuence model.
Findings have implications not only for understand-
ing the proximal correlates of friends depressive symp-
toms, but also for the continued study of antecedents of
a depressogenic attributional style. It will be important
for future work to differentiate factors that uniquely pre-
dict the development of negative cognitions from those
that have implications for attributional style via depressive
symptoms. Although both types of predictors are relevant
to enhance a developmental understanding of attributions,
and both may be useful for prevention efforts, ndings may
vary in the breadth of implications. Findings regarding the
potential effects of friendsdepressive symptoms on ado-
lescents own depressive symptoms have implications for
a range of depression-related consequences (e.g., hope-
lessness, suicidality, substance use), one of which may be
increases in a depressogenic attributional style.
Also of note, ndings from this study provided
stronger support for friends depressive symptoms, as
compared to friendsattributional styles, as predictors of
adolescents own functioning. These results suggest that
aspects of depression other than the communication of
negative explanatory styles might be especially impor-
tant in the interpersonal transmission of depressive symp-
toms. Future work examining specic symptoms or behav-
ioral correlates of depression (e.g., sad affect, lethargy,
anhedonia, hopelessness, reassurance-seeking) may re-
veal unique behaviors that are the most potent agents of
contagion, and may be addressed directly in interventions.
Future research also would benet by addressing
some of the limitations of this investigation. For exam-
ple, although ndings from this study offer important di-
rections for understanding the prevention of depressive
symptoms among adolescents from a community sam-
ple, the study of similar processes in a clinically referred
sample would be useful to elucidate interpersonal factors
that are associated with more severe levels of symptoms.
The use of additional assessment instruments (e.g., struc-
tured clinical interviews) also would be helpful for un-
derstanding symptoms of depressive disorders. Future re-
search examining interpersonal correlates of adolescents
depressive cognitions also might benet from behavioral
observations of social behaviors and the study of mecha-
nisms that may explain the peer-contagion effect. Lastly,
future examination of friends inuence should include
peers both within and outside the school setting to best
capture multiple sources of contagion to which adoles-
cents are exposed.
Overall, the results from this study offer additional
evidence to suggest that adolescents peer relationships
may contribute signicantly to the development of de-
pressive symptoms and cognitions during the transition
to adolescence, particularly among girls. Findings sug-
gest that attempts to modify the cognitions of adolescents
should consider the possibility that these cognitions may
be strongly reinforced within adolescentsclose interper-
sonal relationships. From a prevention perspective, it is
also important to consider the effects of peer homophily
on adolescents adjustment not only for externalizing out-
comes, but also for internalizing symptoms.
ACKNOWLEDGMENTS
This research was supported by grants from NIMH
(R01-MH59766) and the American Foundation of Suicide
Prevention awarded to the second author. Special thanks
are due to Charissa S. L. Cheah, Julie Wargo Aikins,
Valerie Simon, Annie Fairlie, Robin M. Carter, Daryn
David, Carrie Hommel, and Erica Foster for their assis-
tance with data collection, as well as all of the adolescents
and families who participated in this project.
REFERENCES
Abela, J. R. Z. (2001). The hopelessness theory of depression: A test of
the diathesisstress and causal mediation components in third and
36 Stevens and Prinstein
seventh grade children. Journal of Abnormal Child Psychology, 29,
241254.
Abramson, L. Y., Metalsky, G. I., & Alloy, L. B. (1989). Hopelessness
depression: A theory-based subtype of depression. Psychological
Review, 96, 358372.
Baron, R. M., & Kenny, D. A. (1986). The moderatormediator variable
distinction in social psychological research: Conceptual, strategic,
and statistical considerations. Journal of Personality and Social
Psychology, 51, 11731182.
Brown, B. B. (1990). Peer groups and peer cultures. In S. S. Feldman,
& G. R. Elliott (Eds.),At the threshold: The developing adolescent
(pp. 171196). Cambridge: Harvard University Press.
Buhrmester, D., & Furman, W. (1987). The development of companion-
ship and intimacy. Child Development, 58, 11011113.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation
analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ:
Lawrence Erlbaum.
Crosseld, A. G., Alloy, L. B., Gibb, B. E., & Abramson, L. Y. (2002).
The development of depressogenic cognitive styles: The role of
negative childhood life events and parental inferential feedback.
Journal of Cognitive Psychotherapy, 16, 487502.
Cyranowski, J. M., Frank, E., Young, E., & Shear, M. K. (2000). Ado-
lescent onset of the gender difference in lifetime rates of major
depression: A theoretical model. Archives of General Psychiatry,
57, 2127.
Downey, G., & Coyne, J. C. (1990). Children of depressed parents: An
integrative review. Psychological Bulletin, 108, 5076.
Furman, W., & Buhrmester, D. (1992). Age and sex differences in per-
ceptions of networks of personal relationships. Child Development,
63, 103115.
Garber, J., & Flynn, C. (2001). Predictors of depressive cognitions in
young adolescents. Cognitive Therapy and Research, 25, 353
376.
Garber, J., Keiley, M. K., & Martin, N. C. (2002). Developmental trajec-
tories of adolescentsdepressive symptoms: Predictors of change.
Journal of Consulting and Clinical Psychology, 70, 7995.
Gibb, B. E. (2002). Childhood maltreatment and negative cognitive
styles: A quantitative and qualitative review. Clinical Psychology
Review, 22, 223246.
Gibb, B. E., Abramson, L. Y., & Alloy, L. B. (2004). Emotional mal-
treatment from parents, verbal peer victimization, and cognitive
vulnerability to depression. Cognitive Therapy and Research, 28,
121.
Gibb, B. E., Alloy, L. B., Abramson, L. Y., & Marx, B. P. (2003). Child-
hood maltreatment and maltreatment-specic inferences: A test of
Rose and Abramsons (1992) extension of the hopelessness theory.
Cognition & Emotion, 17, 917931.
Gibb, B. E., Alloy, L. B., Abramson, L. Y., Rose, D. T., Whitehouse,
W. G., Donovan, P., et al. (2001). History of childhood maltreat-
ment, negative cognitive styles, and episodes of depression in adult-
hood. Cognitive Therapy and Research, 25, 425446.
Gladstone, T. G., & Kaslow, N. J. (1995). Depression and attributions
in children and adolescents: A meta-analytic review. Journal of
Abnormal Child Psychology, 23, 597606.
Hammen, C., & Brennan, P. A. (2001). Depressed adolescents of de-
pressed and nondepressed mothers: Tests of an interpersonal impair-
ment hypothesis. Journal of Consulting and Clinical Psychology,
69, 284294.
Hammen, C., Burge, D., & Adrian, C. (1991). Timing of mother and
child depression in a longitudinal study of children at risk. Journal
of Consulting and Clinical Psychology, 59, 341345.
Hankin, B. L., & Abramson, L. Y. (2001). Development of gender
differences in depression: An elaborated cognitive vulnerability
transactional stress theory. Psychological Bulletin, 127, 773796.
Hankin, B. L., Abramson, L. Y., Moftt, T. E., Silva, P. A., McGee, R.,
& Angell, K. E. (1998). Development of depression from pread-
olescence to young adulthood: Emerging gender differences in a
10-year longitudinal study. Journal of Abnormal Psychology, 107,
128141.
Hankin, B. L., Abramson, L. Y., & Siler, M. (2001). A prospective test
of the hopelessness theory of depression in adolescents. Cognitive
Therapy & Research, 25, 607632.
Harter, S., Stocker, C., & Robinson, N. S. (1996). The perceived direc-
tionality of the link between approval and self-worth: The liabilities
of a looking glass self-orientation among young adolescents. Jour-
nal of Research on Adolescence, 6, 285308.
Haselager, G. J. T., Hartup, W., Van Lieshout, C. F. M., &
Riksen-Walraven, J. M. (1998). Similarities between friends and
nonfriends in middle childhood. Child Development, 69, 1198
1208.
Hogue, A., & Steinberg, L. (1995). Homophily of internalized distress
in adolescent peer groups. Developmental Psychology, 31, 897
906.
Holmbeck, G. N. (2002). Post-hoc probing of signicant moderational
and mediational effects in studies of pediatric populations. Journal
of Pediatric Psychology, 27, 8796.
Jaenicke, C., Hammen, C., Zupan, B., Hiroto, D., Gordon, D., Adrian, C.,
& Burge, D. (1987). Cognitive vulnerability in children at risk for
depression. Journal of Abnormal Child Psychology, 15, 559572.
Joiner, T. E. (1994). Contagious depression: Existence, specicity to
depressed symptoms, and the role of reassurance-seeking. Journal
of Personality and Social Psychology, 67, 287296.
Joiner, T. E., & Wagner, K. D. (1995). Attribution style and depres-
sion in children and adolescents: A meta-analytic review. Clinical
Psychology Review, 15, 777798.
Kandel, D. B. (1978). Homophily, selection and socialization in adoles-
cent friendships. American Journal of Sociology, 84, 427436.
Kandel, D. B. (1996). The parental and peer contexts of adolescent de-
viance: An algebra of interpersonal inuences. Journal of Drug
Issues, 26, 289316.
Keenan, K., Loeber, R., Zhang, Q., Stouthamer-Loeber, M., & Van
Kammen, W. B. (1995). The inuence of deviant peers on the de-
velopment of boys disruptive and delinquent behavior: A temporal
analysis. Development and Psychopathology, 7, 715726.
Kovacs, M. (1992). Childrens Depression Inventory Manual. North
Tonawanda, New York: Multi-Health Systems, Inc.
Lea, M. (1979). Personality similarity in unreciprocated friendships.
British Journal of Social and Clinical Psychology, 18, 393394.
Lewinsohn, P. M. (1974). A behavioral approach to depression. In R. M.
Friedman & M. M. Katz (Eds.), The psychology of depression: Con-
temporary theory and research (pp. 157185). New York: Wiley.
Nolen-Hoeksema, S., Girgus, J., & Seligman, M. E. (1992). Predictors
and consequences of childhood depressive symptoms: A 5-year
longitudinal study. Journal of Abnormal Psychology, 101, 405422.
Panak, W. F., & Garber, J. (1992). Role of aggression, rejection, and
attributions in the prediction of depression in children. Development
& Psychopathology, 4, 145165.
Parker, J. G., & Asher, S. R. (1993). Friendship and friendship quality in
middle childhood: Links with peer group acceptance and feelings
of loneliness and social dissatisfaction. Developmental Psychology,
29, 611621.
Prinstein, M. J., & Aikins, J. W. (2004). Cognitive moderators of the
longitudinal association between peer rejection and adolescent de-
pressive symptoms. Journal of Abnormal Child Psychology, 32,
147158.
Prinstein, M. J., Boergers, J., & Spirito, A. (2001). Adolescents and
their friendshealth-risk behavior: Factors that alter or add to peer
inuence. Journal of Pediatric Psychology, 26, 287298.
Prinstein, M. J., Meade, C. S., & Cohen, G. L. (2003). Adolescent oral
sex, peer popularity, and perceptions of best friends sexual behav-
ior. Journal of Pediatric Psychology, 28, 243249.
Robbins, C. L., & Hinkley, K. (1989). Social-cognitive processing and
depressive symptoms in children: A comparison of measures. Jour-
nal of Abnormal Child Psychology, 17, 2936.
Rose, A. J. (2002). Co-rumination in the friendship of girls and boys.
Child Development, 73, 18301843.
Rose, D. T., & Abramson, L. Y. (1992). Developmental predictors of
depressive cognitive styles: Research and theory. In D. Cicchetti
Peer Contagion 37
& S. L. Toth (Eds.), Developmental perspectives on depression.
(Vol. 4, pp. 323349). Rochester symposium on developmental psy-
chopathology.
Rosenblatt, A., & Greenberg, J. (1988). Depression and interpersonal
attraction: The role of perceived similarity. Journal of Personality
and Social Psychology, 55, 112119.
Rosenblatt, A., & Greenberg, J. (1991). Examining the world of
the depressed: Do depressed people prefer others who are de-
pressed?Journal of Personality and Social Psychology, 60, 620
629.
Rudolph, K. D. (2002). Gender differences in emotional responses
to interpersonal stress during adolescence. Journal of Adolescent
Health, 30S, 313.
Rudolph, K. D., & Hammen, C. (1999). Age and gender as determi-
nants of stress exposure, generation, and reactions in youngsters: A
transactional perspective. Child Development, 70, 660677.
Sacco, W. P. (1999). A social-cognitive model of interpersonal pro-
cesses in depression. In T. Joiner & J. C. Coyne (Eds.), The in-
teractional nature of depression: Advances in interpersonal ap-
proaches (pp. 329362). Washington, DC: American Psychological
Association.
Saylor, C. F., Finch, A. J., Spirito, A., & Bennett, B. (1984). The Chil-
drens Depression Inventory: A systematic evaluation of psychome-
tric properties. Journal of Consulting & Clinical Psychology, 52,
955967.
Seligman, M. E., Peterson, C., Kaslow, N. J., Tanenbaum, R. L., Alloy,
L. B., & Abramson, L. Y. (1984). Attributional style and depressive
symptoms among children. Journal of Abnormal Psychology, 13,
227231.
Strack, S., & Coyne, J. C. (1983). Social conrmation of dysphoria:
Shared and private reactions to depression. Journal of Personality
and Social Psychology, 44, 798806.
Tambs, K. (1991). Transmission of symptoms of anxiety and depres-
sion in nuclear families. Journal of Affective Disorders, 21, 117
126.
Thompson, M., Kaslow, N. J., Weiss, B., Nolen-Hoeksema, S.
(1998). Childrens Attributional Style QuestionnaireRevised:
Psychometric examination. Psychological Assessment, 10, 166
170.
Urberg, K. A., Degirmencioglu, S. M., & Pilgrim, C. (1997). Close friend
and group inuence on adolescent cigarette smoking and alcohol
use. Developmental Psychology, 33, 834844.
Vitaro, F., Tremblay, R. E., Kerr, M., Pagani, L., & Bukowski, W. M.
(1997). Disruptiveness, friends characteristics, and delinquency in
early adolescence: A test of two competing models of development.
Child Development, 68, 676689.
Wills, T. A., Cleary, S. D. (1999). Peer and adolescent substance
use among 6th9th graders: Latent growth analyses of inu-
ence versus selection mechanisms. Health Psychology, 18, 453
463.
    • So it may be that adolescents with more depressive symptoms have more difficulties with applying social skills to romantic relationships but less so in friendships (especially when the best friend also tries to resolve a conflict with positive problem solving). The results concerning interpersonal effects were in line with prospective studies showing support for depression contagion; that is, levels of depressive symptoms in adolescents and friends were positively related over time (e.g., Prinstein, 2007; Stevens & Prinstein, 2005; Van Zalk et al., 2010b ). This contagion may take place because of maladaptive interactions that depressed adolescents may have (Coyne, 1976 ).
    [Show abstract] [Hide abstract] ABSTRACT: This 6-year longitudinal study examined the relation between 3 conflict management styles (i.e., problem solving, conflict engagement, and compliance) and depressive symptoms in adolescent-best friend relationships. Participants were 479 Dutch adolescents and their best friend who reported annually on depressive symptoms and conflict management styles toward each other. Bidirectional effects between conflict management styles and depressive symptoms were studied both within adolescents (intraindividual) and between adolescent best friends (interpersonal). A positive interpersonal effect of depressive symptoms of one dyad member on depressive symptoms of the other member was found. Similarly, higher positive problem solving and conflict engagement of one dyad member predicted respectively higher problem solving and conflict engagement of the other dyad member. Adolescents who reported more depressive symptoms reported more conflict engagement and compliance over time. In addition, for boys, higher levels of depressive symptoms of one dyad member were related to more problem solving by the other member over time. The current study contributed to the literature by showing that depressive symptoms and conflict management are related constructs in adolescents and that both intrapersonal and interpersonal processes contribute to this relation.
    Full-text · Article · Dec 2016
    • Of interest, the current study did not provide evidence that the mediating effect of co-rumination or the additional moderated effects were stronger for females versus males, or for older versus younger adolescents. These results are consistent with past studies of depression contagion finding nonsignificant age (Giletta et al., 2012; Schwartz-Mette & Rose, 2012) or gender (Schwartz-Mette & Rose, 2012; Stevens & Prinstein, 2005; cf. Giletta et al., 2012 ) moderation effects, and of co-rumination's mediating effect on depression contagion (Schwartz-Mette & Rose, 2012), which also found no significant gender or developmental differences.
    [Show abstract] [Hide abstract] ABSTRACT: Research supports the notion that adolescents’ mental health is impacted by peers via contagion processes. A growing area of interest has been how co-rumination may influence depressive symptoms within friendships. The current study examined particular conditions under which co-rumination is especially likely to facilitate depression contagion. Participants were adolescents (N = 480, 49% female, M age = 14.6 years, 59.5% European American) paired in friendship dyads and assessed over 9 months. Characteristics of the adolescent (personal distress), of the friend (excessive reassurance seeking), and of the friendship (friendship quality) were considered. Moderated mediation analyses indicated that co-rumination facilitated depression contagion only under conditions of adolescents’ high personal distress, friends’ high excessive reassurance seeking, and high positive friendship quality. This research underscores the importance of attending to how and under what conditions depression contagion occurs within friendships in order to support adolescents’ positive social and emotional development.
    Full-text · Article · Sep 2016
    • Recent research concerning peer influence on internalizing problems has indicated that depressive symptoms and attributional styles are contagious within friend dyads and peer groups (Conway et al. 2011; Giletta et al. 2011; Stevens and Prinstein 2005). The influence may occur through discussion of adverse life events and display of negative emotions among group members (Stevens and Prinstein 2005). Withdrawn groups may function as a context socializing internal distress and depressive symptoms as group members discuss their difficulties and frustrations in social interactions and their negative attitudes toward others and life circumstances .
    [Show abstract] [Hide abstract] ABSTRACT: This 1-year longitudinal study examined the effects of membership in socially withdrawn peer groups on children's social and psychological adjustment in a sample of 979 children (417 boys, 562 girls, M age = 11.84 years). Data on children's social and psychological adjustment and problems were collected from peer nominations and self-reports in the fall and spring of a single academic year. Using the Social Cognitive Map, 162 peer groups were identified. Multilevel analyses showed that affiliation with withdrawn groups negatively predicted social competence and school attitude, and positively predicted victimization and depression. The results suggest that affiliation with socially withdrawn groups is a risk factor for the development of social and psychological problems.
    Full-text · Article · Dec 2015
    • However, friends may also impair positive emotional and behavioral development . That is, several studies have shown that youth are often in friendships with similarly aggressive or depressive children (Cairns, Cairns, Nechermann, Gest & Gariepy, 1998; Poulin et al., 1997; Stevens & Prinstein, 2005), often select friends similar on behaviors such as aggression and delinquency (Snyder, Horsch & Childs, 1997; Werner & Crick, 2004) and are also influenced by their friends' problem behaviors (Haynie, 2001; Urberg, 1999 ). In adolescence, these processes become much more problematic as youth spend more time with peers and peers become important, if not the most important, socializing agents (Jang, 1999; Larson, Richards, Moneta, Holmbeck, & Duckett, 1996).
    [Show abstract] [Hide abstract] ABSTRACT: In the current study, associations between individual and friends’ direct and indirect aggression and depressive problems were examined. It was expected that social status would moderate these associations such that low-status preadolescents would be more similar to their unilateral friends with regard to indirect, but not direct, aggression. Furthermore, it was expected that preadolescents’ depressive problems were positively associated to reciprocal friends’ depressive problems, in particular in low-status preadolescents. The hypotheses were tested by studying unilateral and reciprocal friendships in 204 children (Mage¼10.90; SD¼0.78; 44.1% girls). Direct and indirect aggression and depressive problems were assessed via both self- and peer-reports. Social status was assessed via peer-reported rejection (i.e., dislike) and popularity (i.e., who do others want to be associated with). Analyses showed partial support for the hypotheses, showing that individual and unilateral friends’ self-reported indirect and direct aggression were positively associated in preadolescents who were lower on popularity. Moreover, in preadolescents who were more rejected by peers, depressive problems were positively associated to reciprocal friends’ depressive problems. The current study highlights the importance of including social status and distinguishing between unilateral and reciprocal friendships when examining associations between individual and friends’ behavior.
    Article · Sep 2015
    • Swann and colleagues have speculated that pairing with similarly depressed partners serves to verify individuals' self-image (Giesler & Swann, 1999; Swann, 2004). The aggregation of similarly depressed youth has been noted in peer groups and close friendships during early to middle adolescence (Hogue & Steinberg, 1995; Mariano & Harton, 2005; Stevens & Prinstein, 2005 ) and roommate and dating relationships during late adolescence ( Katz, Beach, & Joiner, 1999 ). The current study extends these findings to early romantic development.
    File · Data · May 2015 · The Arts in Psychotherapy
    • By contrast, the positive style is characterized by external, unstable, and specific explanations for bad events (Abramson, Metalsky, & Alloy, 1989;Abramson, Seligman, & Teasdale, 1978;Buchanan & Seligman, 1995). Empirically, cross-sectional and longitudinal studies have indicated a strong association between children's and adolescents' negative explanatory style and a wide range of negative social and emotional outcomes, including depression (Abela & Hankin, 2008;Hankin, Abramson, & Siler, 2001;Joiner, 2000;Prinstein & Aikins, 2004;Rueger & Malecki, 2007;Southall & Roberts, 2002;Stevens & Prinstein, 2005). Similar associations were reported for loneliness, social anxiety (Crick & Ladd, 1993;Toner & Heaven, 2005), as well as low acceptance by peers and peer-rejection (Ames, Ames, & Garrison, 1977;Toner & Munro, 1996).
    [Show abstract] [Hide abstract] ABSTRACT: The current study examined differences in children's and parents’ hope and negative attributional style as a function of children's attachment categorizations derived from children's family drawings in 77 triads of young elementary school age children (Mean age 6.70, range 6-7.5; SD = 53), mothers, and fathers in Israel. In addition, associations between scales in children's family drawings and children's and parents’ hope and negative attributional style were examined. Drawings were coded using Main and Kaplan's (1986) coding system. Both children's and parents’ hopes and attributional style were reflected in the children's drawings. Specifically, the scales reflecting attachment security, such as vitality and pride, were positively correlated with children's hopes and negatively correlated with children's and parents’ negative attributional style. By contrast, the scales reflecting attachment insecurity, such as emotional distance, anger, and pathology, were negatively correlated with participants’ positive attributional style and hope and positively correlated with negative attributional style. Surprisingly, role reversal was positively correlated with mothers’ hope and negatively with fathers’ negative attributional style. Finally, differences in children's hope and negative attributional style, in addition to mothers’ negative attributional style, were found as a function of children's attachment categorizations. The implications for theory and clinical interventions are discussed.
    Full-text · Article · Mar 2015
Show more