Bridget B. Weymouth, Cheryl Buehler, Nan Zhou, and Robert A.

Henson University of
North Carolina at Greensboro

A Meta-Analysis of Parent–Adolescent Conflict:

Disagreement, Hostility, and Youth Maladjustment

Parent–adolescent conflict is a normative char- conflict has been of continued empirical interest
acteristic of adolescence. However, research to developmental, family, and clinical scholars.
findings have been inconsistent, and the rel- The main goal of this study was to organize,
ative contributions of specific dimensions of synthesize, and summarize the findings on
parent–adolescent conflict (disagreement and the association between parent–adolescent
hostility) to youth maladjustment are unknown. conflict and youth maladjustment. We used
This meta-analysis synthesized the literature on meta-analytic techniques to (a) estimate
parent–adolescent conflict and distinguished the association between parent–adolescent
disagreement, hostility, and composite mea- conflict and youth maladjustment and (b)
sures of disagreement and hostility. A multilevel examine determinants of variability in the
model was utilized to analyze 401 effects from 52 effect of parent–adolescent conflict on youth
studies. Results indicate that parent–adolescent maladjustment.
conflict is positively associated with youth Two perspectives have commonly been
maladjustment. The strength of this association used to frame discussions and examinations
varied as a function of youth maladjustment of parent–adolescent conflict. First, previ-
dimensions but not conflict dimensions. The ous research has discussed and examined
association between parent–adolescent con- parent–adolescent conflict as a normative
flict and youth maladjustment also varied and functional characteristic of adoles-
by youth gender and by longitudinal versus cent development. In their meta-analysis of
cross-sectional design. Results suggest that both parent–adolescent conflict, Laursen, Coy, and
disagreement and hostility in parent–adolescent Collins (1998) highlighted the developmentally
relationships have negative effects on youth normative occurrence of parent–adolescent con-
development. flict during adolescence, finding that frequency
of parent–adolescent conflict peaks during
Although research has long diverged from a early adolescence (then decreases linearly), and
“storm and stress” characterization of adoles- that the negative affectivity of these conflicts
cence (Smetana, Campione-Barr, & Metzger, increases from early to middle adolescence.
2006; Steinberg, 1990, 2001), parent–adolescent Research on the topics and reasoning of these
conflicts has suggested that conflicts between
adolescents and parents typically occur over
everyday issues and whether issues are sub-
Department of Human Development and Family Studies,
P.O. Box 26170, Greensboro, NC 27402–6170 ject to (a) adolescent authority, because they
(bbweymou@[Link]). reside within the adolescent-regulated personal
Key Words: Adolescence, conflict, maladjustment, meta– domain, or (b) parent authority, because they
analysis, parent–adolescent conflict. reside within the parent-regulated prudential
Journal of Family Theory & Review 8 (March 2016): 95–112 95
DOI:10.1111/jftr.12126
96 Journal of Family Theory & Review

or conventional domains (Campion-Barr & affect during conflict, despite similar or higher
Smetana, 2010; Smetana, 2005). On the basis frequencies of conflict. Likewise, other studies
of this research, scholars have suggested that have found that conflict frequency is associ-
parent–adolescent conflict facilitates trans- ated with negative youth outcomes only under
formational processes during adolescence in conditions of negative parent–adolescent rela-
which youth attempt to adjust parent–adolescent tionships (Adams & Laursen, 2007). Thus, we
boundaries, renegotiate parental authority, and deconstructed conflict on the basis of how con-
increase their autonomy and independence flicts are expressed between parents and adoles-
(Steinberg, 2001)—all of which are essential cents, distinguishing dyadic, parent–adolescent
elements to typical adolescent growth and disagreement, expressed hostility, and compos-
development (Smetana, 2005). ites of disagreement and hostility. Moreover, we
Second, in addition to research that has examined whether the strength of the effect size
illustrated developmentally normative increases differed across these conflict dimensions.
in conflict, other research has discussed and We defined parent–adolescent disagreement
examined parent–adolescent conflict as a pre- as discussion and/or debate between parents and
dictor of a range of maladaptive emotional adolescents regarding specific issues. Indica-
and behavioral outcomes for youth, includ- tors include opposition and differences in opin-
ing internalizing problems (Caples & Barrera, ions (Adams & Laursen, 2007). In contrast, we
2006), such as depression (Wang, Brinkworth, defined parent–adolescent hostility as specific,
& Eccles, 2013), and externalizing problems overt behavior and expression and/or commu-
(Maggs & Galambos, 1993), such as aggression nication between parents and adolescents that
(Yeh, 2011). The effect estimates for the asso- includes arguing, angry comments, contempt,
ciation between parent–adolescent conflict and yelling, swearing, name-calling, and/or physi-
indicators of youth maladjustment, however, cal aggression (Buehler, 2006; Buehler, Krish-
have been inconsistent in both statistical sig- nakumar, Anthony, Tittsworth, & Stone, 1994).
nificance and direction, with estimates ranging Given that some studies have utilized measures
from –.48 to .75. As such, given variability in in which indicators of disagreement and hostility
the effect estimates of the association between are aggregated in the same scale, we also exam-
parent–adolescent conflict and youth malad- ined composites of disagreement and hostility,
justment, a central objective of this study is to which are conceptualized as parent–adolescent
examine what explains this variability. behaviors that include disagreement as well as
The term parent–adolescent conflict is broad indicators of hostility, including arguing and
and has been used inconsistently in previous contemptuous behavior.
research to describe a multidimensional con- In addition to distinguishing parent–
struct that can include dyadic behaviors ranging adolescent disagreement from hostility, this
from discussions about differences in opinions meta-analysis advances knowledge about the
to heated arguments that might include yelling, association between parent–adolescent conflict
contempt, and aggression. A central tenet of this and youth maladjustment by examining special-
study is that conflict or the frequency of con- ized effects, that is, whether the strength of the
flict does not have inherently negative effects association between parent–adolescent conflict
on youth adjustment, nor is conflict inherently and youth maladjustment differed across four
neutral (Adams & Laursen, 2007). Rather, asso- indicators of youth maladjustment: academic,
ciations between parent–adolescent conflict and externalizing, internalizing, and total problems.
youth maladjustment may depend on how differ- Academic problems were conceptualized as
ences in opinion are expressed between relation- low youth grades or low achievement on stan-
ship partners (Sprey, 1969; Steinberg, 2001). For dardized assessments of school and academic
example, Smetana (1996) utilized cluster analy- achievement. Externalizing problems were
sis to group families on the basis of frequency conceptualized as broad problematic behaviors
of conflict and intensity of negative affect during that are directed outward. Indicators included
conflicts. Results indicated that adolescents from aggression, delinquency, impulsivity, hyperac-
families characterized by higher negative affect tivity, substance abuse, and/or conduct problems
were more detached from parents and had lower (Achenbach & Edelbrock, 1978; Buehler et al.,
levels of academic performance than adoles- 1997). Internalizing problems were conceptual-
cents from families with lower levels of negative ized as broad problematic behaviors involving
Parent–Adolescent Conflict 97

issues with self (rather than others). Indicators dysregulation. Similarly, according to social
included depression, anxiety, low self-esteem, learning theory, parent–adolescent interactions
social withdrawal, and/or general psycholog- serve as models for youths’ future behavior
ical distress (Achenbach & Edelbrock, 1978; in other social contexts, such that youth may
Buehler et al., 1997). Total problems were mimic their interactions with their parents
conceptualized as a global indicator of external- in other social interactions (Bandura, 1973;
izing and internalizing behaviors (Achenbach Goldhaber, 2000). Adolescents in parent–youth
& Edelbrock, 1978) in which a combination dyads that engage in more hostile behaviors, and
of broad indicators or a combination of their conversely, dyads that are able to discuss dif-
more specific indicators—including aggression, ferences more constructively, will likely display
depression, anxiety, conduct problems and/or similar behaviors in other social contexts.
delinquency, self-esteem, and/or substance In line with attachment and social learning
use—were measured by one scale. theories, several studies have found significant
associations between parent–adolescent conflict
and a multitude of negative youth outcomes. In
Theoretical and Empirical Framework studies of parent–adolescent hostility and broad
On the basis of a developmental psychopathol- indicators of youth maladjustment, researchers
ogy perspective (Cicchetti & Rogosch, 2002; have found that higher parent–adolescent hos-
Cummings, Davies, & Campbell, 2000), we uti- tility is associated with greater internalizing
lized several theoretical frameworks to guide our (Crean, 2008), greater externalizing (Barrera
meta-analysis. Theoretically, disagreement with & Stice, 1998), and higher total problems
parents may reflect important social-cognitive (Buehler & Gerard, 2002). Significant asso-
processes. During adolescence, youth develop ciations also have been substantiated among
more advanced cognitive abilities that facilitate parent–adolescent hostility and more specific
youth reflection on parents’ social conventions indicators of youth maladjustment, including
and the boundaries that divide the issues that are higher depression (El-Sheikh & Elmore-Staton,
subject to parents’ authority versus adolescents’ 2004), anxiety (Pasch et al., 2006), aggression
personal jurisdictions (Smetana, 1996; Stein- (Steeger & Gondoli, 2013), conduct prob-
berg & Silk, 2002). Moreover, parent–adolescent lems (Simons-Morton, Chen, Hand, & Haynie,
disagreement may promote adolescent individ- 2008), and lower self-esteem (Kulhberg, Pena,
uation processes (Blos, 1979). From a family & Zayas, 2010).
systems perspective, disagreement (as opposed Contrary to expectations from family
to hostility) might be necessary for the estab- systems theory, other research that has exam-
lishment of new boundaries within family sys- ined only disagreement also has found that
tems and can facilitate typical, developmentally parent–adolescent disagreement is associated
appropriate youth differentiation from parents with negative youth outcomes. Similar to
and a balance between individuality and con- findings related to parent–adolescent hostil-
nectedness in the parent–adolescent relation- ity, studies have found that parent–adolescent
ship (Bowen, 1978; Grotevant & Cooper, 1985; disagreement is associated significantly with
Smetana, 2005; Smetana, Abernethy, & Har- higher youth depression (Bámaca-Colbert,
ris, 2000). As such, disagreement may result in Umaña-Taylor, & Gayles, 2012) and delin-
less negative outcomes for youth development quency (Bradford, Vaughn, & Barber, 2008).
(Gavazzi & Sabatelli, 1990) than hostility and Likewise, research that has utilized com-
composites of disagreement and hostility. posite measures of parent–adolescent conflict
Parent–adolescent disagreements that are that include both disagreement and hostility
characterized by fighting, aggression, and con- has found that parent–adolescent conflict is
tempt, however, could damage both trust and associated with higher internalizing (Caples &
secure attachment bonds between parents and Barrera, 2006), externalizing (Maggs & Galam-
their children (Bowlby, 1973), thereby altering bos, 1993), and total problems (McKinney &
youths’ internal working models of self and Renk, 2011). More specific associations also
relationships. From this perspective, hostile have been substantiated between composite
parent–adolescent relationships may contribute measures and higher depressive symptoms
to adolescents’ internal working models that are (Wang et al., 2013), aggression (Yeh, 2011),
characterized by more emotional and behavioral and conduct problems (Wang et al., 2013). In
98 Journal of Family Theory & Review

summary, a review of the literature suggests that Parent gender. Similar to youth gender, incon-
disagreement, hostility, and composites of dis- sistencies also have been found regarding
agreement and hostility may each be associated parent gender. Some studies have found that
with youth maladjustment. father–adolescent, but not mother–adolescent,
We also examined several effect- and hostility is associated positively with youth
study-level characteristics that may moder- externalizing problems (Richmond & Stocker,
ate the association between parent–adolescent 2006) and total problem behaviors (Hakvoort,
conflict and youth maladjustment and possi- Bos, van Balen, & Hermanns, 2010). Like-
bly help clarify research findings. Effect-level wise, longitudinal studies have suggested that
moderators included parent and youth gender, father–adolescent, but not mother–adolescent,
measurement source and/or informant, effect hostility is associated with greater subsequent
design, and effect quality. Due to an insuffi- emotional distress (Summers, Forehand, Armis-
cient number of studies that utilized samples of tead, & Tannenbaum, 1998) and decreased youth
LBGT parents and/or provided information on well-being (Shek, 1998). In contrast, other find-
parental sexual orientation, the latter was not ings have suggested that mother–adolescent,
coded as an effect-level moderator. Study-level but not father–adolescent, hostility is asso-
moderators included publication year and type, ciated with greater youth emotional distress
average youth age in sample, predominant (Manne & Miller, 1998) and increases in
race in the sample, and family structure and/or youth antisocial behavior over time (Summers
composition. et al., 1998). Other research has suggested
that mothers spend more time with their chil-
dren than do fathers (Bronstein, 1984) and
Effect-Level Moderators that mother–adolescent relationships are more
intimate than father–adolescent relationships
Youth gender. We expected that the strength (Steinberg & Silk, 2002). As such, in the context
of the association between parent–adolescent of less intimacy and contact with fathers, it is
conflict and youth maladjustment might depend possible that father–adolescent conflict is more
on youth gender. The ways girls and boys salient and has greater implications for youth
engage in relationships differ, such that girls maladjustment (Hakvoort et al., 2010) than
tend to be more oriented than boys by com- does mother–adolescent conflict. This study
munion beliefs, characterized by appeasement, examined parent gender as a moderator of the
closeness, and harmony maintenance behaviors association between parent–adolescent conflict
(Davies & Lindsay, 2004). Research, however, and youth maladjustment to determine whether
has been inconsistent regarding how youth any systematic pattern was evident across this
gender might moderate associations between body of research; however, we did not formulate
parent–adolescent conflict and youth maladjust- specific hypotheses given that previous findings
ment. Stevens, Vollebergh, Pels, and Crijnen are inconsistent.
(2007) found that parent–adolescent hostil-
ity was associated with greater internalizing Measurement source and/or informant. The
and externalizing symptoms among daughters strength of effect sizes also might depend on
but not sons. Likewise, results from Stewart, the informant of the data who was used to
Lam, Betson, and Chung (1999) indicated that estimate associations, given that scholars have
parent–adolescent disagreement was associated suggested that various family members have
positively with suicidal ideation among daugh- different perspectives on parent–adolescent
ters but not sons. In contrast, other findings conflict (Steinberg, 2001). For example,
have indicated that parent–adolescent hostility research has suggested that youth reports of
is associated with greater academic and inter- parent–adolescent conflict tend to be more
nalizing problems among sons but not daughters closely aligned with observational reports than
(Tesser, Forehand, Brody, & Long, 1989). Given with parent reports (Gonzales, Cauce, & Mason,
the inconsistent findings of previous research, 1996). Moreover, Steeger and Gondoli (2013)
we did not formulate specific hypotheses about found that the significance of associations
how youth gender might moderate the associ- between mother–adolescent conflict and youth
ation between parent–adolescent conflict and depressive symptoms and aggression differed
youth maladjustment. between youth and mother reports of these
Parent–Adolescent Conflict 99

variables. As such, we examined the informant conflict and youth depressive symptoms and
of parent–adolescent conflict and the informant aggression among youth from sixth to seventh
of youth maladjustment as moderators of the grades, but not among youth from seventh to
association between parent–adolescent conflict eighth grades (Steeger & Gondoli, 2013). In con-
and youth maladjustment; however, we did not trast, other studies have found significant associ-
formulate specific hypotheses, given the lack ations across adolescence (Briere, Archambault,
of previous research comparing the strength of & Janosz, 2013).
associations between parent–adolescent conflict
and youth maladjustment as a function of who Predominant sample race. Scholars have sug-
reported on these variables. gested that there might be variation across
racial and ethnic groups in their experi-
Effect design and quality. We also expected that ences of parent–adolescent conflict (Dixon,
the strength of effect sizes may differ according Graber, & Brooks-Gunn, 2008), because
to their design (longitudinal vs. cross-sectional) parent–adolescent conflict might be contex-
and their methodological rigor. Less method- tualized by family expectations and values
ologically rigorous studies are likely to overesti- that differ across racial and ethnic groups. For
mate effect sizes (Jüni, Altman, & Egger, 2001), example, studies suggest that African American
and results from meta-analyses can be biased if and Latino families place greater emphasis,
various indicators of method quality (e.g., longi- overall, on obedience, respect, and parental
tudinal vs. cross-sectional, multiple reporters vs. authority than do European American families
single reporters) are not taken into account. (García Coll, Meyer, & Brillon, 1995; Smetana,
Crean, & Daddis, 2002). In these familial
Study-Level Moderators contexts, adolescent engagement in conflict
Publication year and type. We also expected with parents is not as accepted or normative
that the strength of the association between (Fuligni, 1998). Therefore, parent–adolescent
parent–adolescent conflict and youth maladjust- conflict might be associated more strongly with
ment might differ on the basis of publication youth maladjustment among African American,
year and type of publication (published vs. the- Latino, and Asian American youth than among
ses or dissertation). Studies, and the research European American youth.
methods used in studies, may improve over time
as a result of a natural accumulation of knowl- Family structure and/or composition. Divorce,
edge and advancement in research methods. As single parenthood, and repartnering bring
such, older studies might be more susceptible multiple transitions to the parent–adolescent
to Type I and Type II errors, which may lead relationship that may inherently alter processes
to underestimation or inflation of effect esti- in family systems (Demo & Fine, 2010), and
mates. Moreover, published empirical journal thus the extent to which parent–adolescent
articles typically undergo a stronger, indepen- conflict affects youth adjustment in com-
dent peer-review process than do dissertations or parison to first-marriage, two-parent fami-
theses. lies. For example, parents and adolescents
from first-marriage, two-parent families
Mean age of youth. We also expected that the may have more opportunities to engage in
strength of effect estimates may depend on mother–adolescent and father–adolescent con-
the age of youth in the sample. Given norma- flict (Smetana, Yau, Restrepo, & Braeges,
tive peaks in parent–adolescent conflict during 1991), and thus associations with youth
early adolescence and increases in the inten- maladjustment might be stronger than asso-
sity of conflict from early to middle adolescence ciations among parent–adolescent dyads from
(Laursen et al., 1998), early to middle adoles- other family structures. However, divorce or
cence may represent a developmental window repartnering may place additional strains on
in which conflict may have particularly impor- parent–adolescent relationships in comparison
tant implications for transformations in rela- to first-marriage, two-parent families (Demo
tionships and youth maladjustment (Steinberg, & Fine, 2010). As such, we expected that the
2001). Research, however, has provided incon- strength of effect estimates might depend on the
sistent findings. Some studies have found a sig- family structure and/or composition of study
nificant association between mother–adolescent samples, but a lack of research on the effects of
100 Journal of Family Theory & Review

family structure and/or composition preclude (i.e., externalizing) dysregulation. Furthermore,

formal hypotheses. social learning theory suggests that greater
youth engagement in coercive behavior with
their parents will likely generalize to their
Summary of Study Objectives behavior in other contexts (Cummings et al.,
and Hypotheses 2000), such that youth will more likely engage
In summary, the goal of this meta-analysis in more coercive behaviors (i.e., externalizing).
was to organize, synthesize, and summarize Thus, we expected that there is significant
available findings on the association between variation in the strength of the association
parent–adolescent conflict and youth malad- between parent–adolescent conflict and youth
justment. Guided by previous research and maladjustment when youth maladjustment
theory, we hypothesized that parent–adolescent is disaggregated into youth academic, exter-
conflict is positively associated with youth nalizing, internalizing, and total problems.
maladjustment. However, we also expected that Specifically, we hypothesized that associations
there is significant variability across studies are strongest between parent–adolescent conflict
in effect estimates. Using expectations from and youth externalizing behaviors, followed by
developmental and family systems theories, associations with youth total, internalizing, and
we hypothesized that there is significant vari- academic problems.
ation in the strength of the association when We also examined several effect- and
parent–adolescent conflict is disaggregated study-level moderators. We expected that
into disagreement, hostility, and composites the strength of the association between parent–
of disagreement and hostility. Specifically, we adolescent conflict and youth maladjustment
expected that the positive association between might differ on the basis of effect-level char-
parent–adolescent hostility and youth malad- acteristics including parent and youth gender,
justment is stronger than associations among measurement source and/or informant, effect
parent–adolescent disagreement, composite design, and effect quality, as well as study-level
measures of disagreement and hostility, and characteristics including publication year and
youth maladjustment. Given theoretical per- type, average youth age in sample, predominant
spectives that suggest that disagreement may race in the sample, and family structure and/or
facilitate developmentally normative increases composition. However, conflicting findings and
in youth individuation (Blos, 1979) and a bal- a lack of previous research on these charac-
ance of individuality and connectedness in teristics overall precluded our formulation of
parent–adolescent relationships (Grotevant & specific hypotheses.
Cooper, 1985), some of these functional com-
ponents of disagreement may minimize some of Method
the hypothesized negative influences of hostility
on youth maladjustment. However, we expected Literature Search and Study Selection
that the association between composite mea- Studies were retrieved primarily using searches
sures of disagreement and hostility and youth of the online databases PsycINFO, EBSCOhost,
maladjustment is stronger than the associations PubMed, and Dissertation Abstracts Interna-
between parent–adolescent disagreement and tional. Conflict subjects and/or keywords (i.e.,
youth maladjustment. conflict, hostil* , argu* , disagree* , family con-
In addition to disaggregating parent– flict, parent–child communication, parent–child
adolescent conflict, we also examined four relations, mother–child relations) were used in
specific indicators of youth maladjustment: combination with youth maladjustment subjects
academic, externalizing, internalizing, and and/or keywords (i.e., adjustment, maladjust-
total problems. Attachment theory (Bowlby, ment, behavior problems, well-being, devel-
1973) suggests that interactions with parents opment) to identify studies. Keywords and/or
will affect youths’ internal working models of subjects specifying parent–child or parent–
themselves and their relationships. As such, adolescent conflict (i.e., parent* , child* , adol* )
engagement in more parent–adolescent conflict also were used. References of identified articles
may change youths’ internal working models were searched to locate additional studies. Pre-
such that their models are characterized by more cursory examination of study titles and abstracts
emotional (i.e., internalizing) and behavioral identified in the search resulted in 135 studies
Parent–Adolescent Conflict 101

that were examined more closely with our Dependent variable characteristics. For each
selection criteria. This meta-analysis included effect, coded characteristics included youth mal-
both available theses and dissertations and adjustment (academic, externalizing, internaliz-
journal publications. Theses and dissertations ing, or total problems), youth maladjustment
and journal publications with similar titles and informant (youth, parent, observer, teacher, or
authors were examined more closely to make peer), and whether youth maladjustment was
sure that duplicate effect estimates were not constructed as a manifest or latent variable.
included. We included only theses and disserta-
tions if they were not, to our knowledge, later Effect design characteristics. For each effect,
published in an academic journal. we also coded whether the effect was longitu-
Only studies that met the following selection dinal or cross-sectional, whether the effect was
criteria were included for coding: (a) the study adjusted for covariates, whether the effect was
was published in English; (b) the study included statistically significant, and the utilized p-value.
a sample of youth with a mean age between Effect quality. We assessed the methodological
10 and 18 years old; (c) the study included a quality of each effect estimate using the fol-
measure of parent–adolescent conflict consis- lowing criteria (Buehler et al., 1997): (a) sam-
tent with our conceptual definitions of disagree- ple quality (.5 points for a study based on 100
ment, hostility, or a composite of disagreement or more participants and .5 points for a random
and hostility; (d) the study included a mea- or stratified sample); (b) longitudinal design (1
sure of youth maladjustment consistent with our point if the effect was based on a longitudinal
conceptual definitions; (e) the study included a design); (c) measurement quality (.5 points if the
quantitative estimate of the association between independent or dependent variable was assessed
parent–adolescent conflict and an indicator of using multiple methods, .25 points if the inde-
youth maladjustment; and (f) the study included pendent variable was assessed using multiple
data needed to calculate and estimate a correla- informants, and .25 points if the dependent vari-
tion coefficient (the effect size reported in the able was assessed using multiple informants);
current study). The selection procedure resulted and (d) analysis quality (1 point if the effect esti-
in 52 studies that met inclusion criteria, resulting mate was based on structural equation model-
in a sample of 401 effects. ing, or .5 points if the effect estimate was based
on a partial correlation or beta coefficient that
included controls). The index of effect quality
Coding and Data Set Preparation ranged from 0 to 4.
The primary effect estimate used in this study Study and sample characteristics. For each
was Pearson’s correlation coefficient (r). Beta study, coded study characteristics included
coefficients were adjusted using the formula publication year and type (published empirical
recommended by Peterson and Brown (2005) article, book chapter, or thesis or dissertation),
(Proulx, Helms, & Buehler, 2007). The primary sampling procedures (random, purposive, or
information coded for each effect included convenient), and study design (longitudinal,
independent variable characteristics, depen- cross-sectional, or mixed). Characteristics of the
dent variable characteristics, effect design sample that were coded included total sample
characteristics, and an assessment of effect size, mean age of youth in the sample, family
methodological quality. We also coded for study structure and/or composition of the sample, and
and sample characteristics. racial composition of the sample.
Interrater reliability. Effects were coded by
Independent variable characteristics. For each the first author. To assess coding reliability, the
effect, coded characteristics included the type of second and third authors coded a combined,
conflict (disagreement, hostility, or a compos- random 25% of effects (n = 104). Interrater
ite of disagreement and hostility), parent gen- reliability was 98% agreement (Cohen’s kappa
der (mother, father, or male or female parent), range = .63–1.00). Coding disagreements were
conflict informant (youth, parent, or observer), resolved through discussion, and any resulting
youth gender, and whether conflict was con- changes to coding guidelines were applied to
structed as a manifest or latent variable. the entire data set.
102 Journal of Family Theory & Review

Statistical Analyses we conducted sensitivity analyses, which com-

pared the results from the multilevel analyses
Effect-size calculations. Of the 52 studies
with those generated using an alternative esti-
that included 401 effect sizes, 291 of the effect
mator, the sandwich estimator. The sandwich
estimates were Pearson’s correlation coefficients
method is a robust estimation method that
(r) and 110 were standardized coefficients (𝛽).
deals with the estimation of dependence (Yuan
Standardized coefficients were transformed into
& Bentler, 2002). It computes the estimated
r (Peterson & Brown, 2005). Before pooling
variance–covariance matrix of fixed-effect
effect estimates, correlations were transformed parameters by using the asymptotically consis-
using Fisher’s Zr transformation (Rosenthal, tent estimator. We also coded the number of
1991). Pooled Zrs were transformed back into r effect estimates for a given study and included
for analyses and reporting. this number of estimates per study as a modera-
tor to examine whether the overall weighted and
Meta-analytic integration and analytic unweighted effect size differed across studies
procedures. A three-level random-effects model with differing numbers of effect estimates.
was used to address the issue of dependent effect
estimates (i.e., several effect estimates in a given Publication bias. We used two methods to
study). Multilevel models can be more appropri- assess possible publication bias. First, we used a
ate for incorporating multiple effect estimates funnel plot (Torgerson, 2006). This plot depicts
within the same study than approaches used the distribution of sample size (standard error)
in past meta-analyses, such as “shifting unit as a function of effect size. We used the average
of analysis” (Kuppens, Laurent, Heyvaert, Fisher’s Z transformed study effect estimates
& Onghena, 2013; Marsh, Bornmann, Mutz, when creating the funnel plot to ensure indepen-
Daniel, & O’Mara, 2009), and do not require dence. In the absence of publication bias, the
independence of effect estimates, as compared plot should look like a funnel, with studies dis-
to traditional random-effects models (Van den tributed symmetrically around the mean Fisher’s
Noortgate & Onghena, 2003). The sampling Z transformed effect size. Second, published
variation for each effect estimate (Level 1), empirical journal articles typically undergo a
variation between effect estimates within a stronger, independent peer-review process than
study (Level 2), and variation between studies do dissertations or theses. Thus, we examined
(Level 3) were modeled. The overall estimate whether the effect size varied by publication
of the association between parent–adolescent status (published vs. unpublished).
conflict and youth maladjustment was obtained
in the unconditional model and subsequent Results
moderators at both the effect level (i.e., Level
The description of individual studies included in
2) and the study level (i.e., Level 3) were fitted.
this meta-analysis is found in Table A1 (see the
Separate models for each moderator variable
online appendix, at [Link]). A
were fitted to avoid inflated Type II error rates
summary of results is in Table 1. Reported coef-
(Raudenbush & Bryk, 2002). Parameters were
ficients are the average, transformed Fisher’s Z
estimated using the restricted maximum likeli-
score, which corresponds to r.
hood procedure implemented in the SAS PROC
MIXED procedure (Littell, Milliken, Stroup,
Wolfinger, & Schabenberger, 2006). Effect Parent–Adolescent Conflict and Youth
estimates were weighted by the inverse of the Maladjustment
sampling variance, and a general Satterthwaite Results indicated that the weighted, average,
approximation was used for the denominator transformed Fisher’s Z score, which corresponds
degrees of freedom for the tests of regression to r, was .27 (p < .001; 95% CI [−.04, .59]).
coefficients. Ninety percent of effect sizes were in the hypoth-
esized positive direction. Moreover, effect sizes
Sensitivity analyses. Although we utilized a differed significantly between (𝜎v2 = .03, p < .01)
three-level random-effects model to address and within studies (𝜎u2 = .01, p < .01).
dependent effect estimates, biased standard Contrary to our hypotheses, the omnibus
errors still may have resulted from insufficient test indicated that the strength of the associa-
modeling of the complex dependence. Thus, tion did not differ significantly across conflict
Parent–Adolescent Conflict 103

dimensions (F(2, 133) = 2.23, p > .05). Also the association between parent–adolescent con-
contrary to our expectations, effect sizes were flict and youth maladjustment did not differ
larger for composite measures of conflict between published and unpublished studies
(r = .31, p < .001, 95% CI [−.06, .68]) than (t = 1.82, p > .05) (see Table 1). As such,
for disagreement (r = .22, p < .001, 95% CI there seems to be minimal estimation error
[−.15, .59]) or hostility (r = .23, p < .001, 95% attributable to publication bias.
CI [−.14, .60]), but the differences were not
statistically significant.
Consistent with our overall hypothesis, the Effect-Level Moderators
omnibus test indicated that the strength of the Several effect-level characteristics moderated
association differed significantly across youth the association between parent–adolescent
maladjustment dimensions (F(3, 29.7) = 7.81, conflict and youth maladjustment. The effect
p < .001). However, contrary to our specialized size differed significantly across youth gender
expectations, follow-up tests indicated that the (F(2, 398) = 26.57, p < .001). Follow-up tests
effect size for externalizing problems (r = .29, indicated that the effect size for estimates based
p < .001, 95% CI [−.03, .61]) was not signif- on samples of both boys and/or girls (r = .28,
icantly different from the effect size for total p < .001, 95% CI [−.04, .60]) was significantly
problems (r = .36, p < .001, 95% CI [−.06, .78]) larger than the effect size for estimates based
or internalizing problems (r = .26, p < .001, 95% on samples of only girls (r = .23, p < .001, 95%
CI [−.06, .58]). As expected, the effect size for CI [−.10, .56]) or only boys (r = .18, p < .001,
total problems was significantly greater than the 95% CI [−.16, .52]). Also, the effect size was
effect size for internalizing problems. Also as significantly larger for samples of only girls
expected, the effect size was the lowest for youth than for only boys. In contrast, the effect size
academic problems (r = .11, p < .001, 95% CI did not differ significantly across parent gender
[−.31, .53]). (F(2, 90.7) = 1.95, p > .05).
The effect size also differed across the mea-
surement source and/or informant. Specifically,
Sensitivity Analysis
the effect size differed significantly across the
Using the sandwich estimator method, results conflict informant (F(3, 251) = 4.30, p < .001).
of the sensitivity analyses indicated that the Follow-up tests indicated that significant differ-
estimates and standard errors for the overall ences were substantiated only between youth
weighted and unweighted effect sizes were report and parent report of conflict. The effect
similar to those from multilevel analyses (see size was larger for youth report of conflict
Table A2). In addition, five studies contributed (r = .27, p < .001, 95% CI [−.06, .60]) than for
more than 20 effect estimates and seven studies parent report of conflict (r = .23, p < .001, 95%
contributed more than 10 but less than 20 effect CI [−.11, .57]). Additionally, the effect size dif-
estimates to the sample of effects because they fered significantly across the informant of youth
examined multiple maladjustment variables or maladjustment (F(2, 366) = 11.50, p < .001).
multiple conflict dimensions, and/or they used Follow-up tests indicated that the effect size
multiple informants of parent–child conflict for composite reports (i.e., a combination of
and/or maladjustment. Results showed that the multiple reporters) of youth maladjustment
estimated weighted and unweighted effect sizes (r = .37, p < .001, 95% CI [−.21, .81]) was
did not differ by the number of effect estimates significantly larger than the effect sizes for
a study contributed to the total sample of effects youth (r = .26, p < .001, 95% CI [−.06, .60])
used in the analyses. This demonstrated that and parent (r = .28, p < .001, 95% CI [−.11,
our reported estimated effect size is robust with .57]) reports of youth maladjustment.
regard to estimation procedures. Moreover, the effect design and methodologi-
cal quality significantly moderated associations.
The effect size differed significantly by effect
Publication Bias
design (t = 10.72, p < .001), such that the
The funnel plot (see Figure A1) showed an effect size for longitudinal effects (r = .21,
approximately symmetrical pattern around the p < .001, 95% CI [−.11, .53]) was larger than
mean, Fisher’s Z transformed effect estimate, cross-sectional effects (r = .12, p < .001, 95%
suggesting minimal publication bias. Moreover, CI [−.21, .45]). The effect size also differed
104 Journal of Family Theory & Review

Table 1. Summary of Multilevel Meta-Analytic Results of Parent–Adolescent Conflict Effect Estimates

Variable K Estimate SE Test Statistic

Overall mean
Weighted 401 r = .27*** .026 t = 10.51***
Unweighted 401 r = .24*** .050 t = 4.88***
Effect-level moderators
Conflict dimensions F(2, 133) = 2.23
Hostility 125 r = .23*** .035
Disagreement 67 r = .22*** .036
Composite 209 r = .31*** .035
Maladjustment dimensions F(3, 29.7) = 7.81***
Externalizing problems 140 r = .29ab *** .027
Internalizing problems 206 r = .26b *** .026
Total problem behaviors 27 r = .36a *** .045
Academic problems 21 r = .11c *** .047
Youth gender F(2, 398) = 26.57***
Boys 18 r = .18c *** .030
Girls 59 r = .23b *** .029
Both boys and girls 324 r = .28a *** .027
Parent gender F(2, 90.7) = 1.95
Mother 149 r = .23*** .035
Father 80 r = .24*** .035
Parent 172 r = .30*** .038
Conflict informant F(3, 251) = 4.30**
Youth report 252 r = .27a *** .028
Observer report 30 r = .26ab *** .071
Composite 47 r = .30ab *** .068
Parent report 72 r = .23b *** .030
Maladjustment informant F (2, 366) = 11.50***
Youth report 283 r = .26b *** .025
Composite 27 r = .37a *** .032
Parent report 59 r = .28b *** .028
Effect design t = 10.72***
Cross-sectional design 294 r = .12b *** .029
Longitudinal design 107 r = .21a *** .026
Methodological rigor 401 𝛽 = −.01 .006 t = −2.30*
Study-level moderators
Publication type t =1.82
Journal article 308 r = .28*** .028
Thesis or dissertation 85 r = .16* .064
Publication year 401 𝛽 = .001 .004 t = 0.25
Youth age 401 𝛽 = .002 .001 t = 1.68
Predominant race of sample F(4, 41.6) = 0.84
White 155 r = .26*** .041
Black 25 r = .20* .098
Hispanic 47 r = .30*** .071
Asian 84 r = .20** .064
Combination 80 r = .34*** .070
Family structure and/or composition of sample t = 0.54
Two-parent families 92 r = .30*** .052
Combination 285 r = .26*** .031
Note. Superscripts a, b, and c are used to illustrate differences between subgroups: Subgroups with different superscripts
represent significant differences. Subgroups with the same superscript represent nonsignificant differences. Subgroups with
two superscripts represent nonsignificant differences between multiple subgroups.
*p < .05. **p < .01. ***p < .001.
Parent–Adolescent Conflict 105

significantly by the methodological qual- examined the contributions of adolescent age

ity of effects. The effect size decreased by and pubertal maturation to parent–adolescent
−0.01 if methodological quality increased by conflict. Contrasting parent–adolescent conflict
a standard deviation (b = −.01, p < .05, 95% during early, middle, and late adolescence,
CI [−.16, .14]). Laursen et al. found weighted effect sizes
ranging from –.31 to .06. Other meta-analyses
from parallel research areas have focused
Study-Level Moderators on the association between interparental
There was no evidence to suggest moderation conflict and parenting behaviors, including
by any of the examined study-level character- parent–adolescent conflict (Krishnakumar &
istics. Specifically, the effect size did not vary Buehler, 2000), as well as youth maladjustment
significantly as a function of the study publi- and interparental conflict (Buehler et al., 1997),
cation year (b = .002, p > .05, 95% CI [−.12, and found weighted effect sizes of –.62 and
.12]), average youth age (b = .002, p > .05, .32, respectively. Thus, the effect size for this
95% CI [−.06, .06]), predominant race of examination is in line with other meta-analyses
study samples (F (4, 41.60) = 0.84, p > .05), or that have included parent–adolescent conflict
family structure and/or composition (t = 0.54, or youth maladjustment as a central variable.
p > .05). However, in comparison to previous research, a
major strength of this study was the use of mul-
tilevel modeling to analyze the effects. The use
Discussion of multilevel modeling decreased dependency
Effect estimates from previous research on threats and allowed for the use of multiple-effect
the associations between parent–adolescent estimates that were located within single stud-
conflict and indicators of youth maladjustment ies. As such, this study did not rely on choosing
have been inconsistent, ranging from –.48 single effects from studies or averaging across
to .75. We posited that conflict is not inher- effect estimates; thus, the effect size obtained in
ently neutral or negative (Adams & Laursen, this study is more comprehensive and represents
2007) and that unspecified conceptualizations a less biased estimation of the total body of
of parent–adolescent conflict and of youth research.
maladjustment may have obscured important Findings also indicated significant variance
distinctions that need to be considered with around the effect size. Contrary to our hypoth-
regard to the implications of parent–adolescent esis, however, this significant variation was
conflict for youth development. Thus, this not predicted by the specific dimensions of
meta-analysis synthesized the available findings parent–adolescent conflict we addressed. Dis-
on the association between parent–adolescent agreement was as strongly associated with youth
conflict and youth maladjustment. Moreover, we maladjustment, as were hostility and composites
differentiated among disagreement, expressed of disagreement and hostility. Although previ-
hostility, and composites of disagreement and ous research has been inconsistent, the current
hostility to examine their relative contributions results are congruent with those studies that have
to youth maladjustment. We also examined found that parent–adolescent disagreement is
several effect- and study-level moderators. significantly associated with greater depression
To our knowledge, this is the first (Bámaca-Colbert et al., 2012) and delinquency
meta-analysis that has examined the associ- (Bradford et al., 2008). The findings support
ation between parent–adolescent conflict and propositions from attachment (Bowlby, 1973)
youth maladjustment. The results, drawn from and social learning (Bandura, 1973) theories,
52 studies and 401 effect estimates, confirmed such that heightened parent–adolescent con-
that parent–adolescent conflict is associated flict may create a context of negativity that
positively with youth maladjustment. The effect influences how youth respond emotionally
size for the association was .27, and 90% of the and behaviorally or replicate negative inter-
effect estimates indicated a positive association actions in other social contexts. According to
between parent–adolescent conflict and youth a family systems perspective, disagreement
maladjustment. Only one other meta-analysis, might be necessary for the establishment of
to our knowledge, has focused primarily on new boundaries within family systems and can
parent–adolescent conflict. Laursen et al. (1998) promote typical youth individuation processes
106 Journal of Family Theory & Review

(Blos, 1979). However, the results also suggest and youth maladjustment differed significantly
that parent–adolescent disagreement may not across youth maladjustment dimensions. Con-
be inherently neutral (Adams & Laursen, 2007) trary to our hypothesis, the association was
given its contribution to negative outcomes for not strongest for youth externalizing problems.
adolescents. Instead, associations were significantly stronger
Research by Smetana (1996) and the use of for youth total problems than for youth inter-
the social domain perspective (Smetana, 2005) nalizing problems; associations were lowest
may offer directions for future research. The for youth academic problems. Consistent with
topic of conflict and the level of discrepancies attachment theory, these results suggest that
between parents’ and adolescents’ perceptions the effects of parent–adolescent conflict do
regarding the domain of this topic, may have not canalize into one maladjustment domain.
important implications for youth maladjustment Our findings are congruent with research that
(Smetana, 1996). Additionally, from a family has found that parent–adolescent conflict is
systems perspective, future research might con- associated with a range of outcomes for youth
sider the hierarchical nature of power within including higher internalizing (Caples & Bar-
some families (Faber, 2002) as an influence rera, 2006), externalizing, and total problems
on the nonsignificant differences between con- (McKinney & Renk, 2011). The results also
flict dimensions in the association between indicate that parent–adolescent conflict is most
parent–adolescent conflict and youth maladjust- strongly associated with youth maladjustment
ment. Family systems theories focus both on when it manifests in a combination of behav-
hierarchical and on more symmetrical structures ioral issues, such as aggression, general conduct
in families (Minuchin, 1974). Adolescence problems, or delinquency, and mental health
is a developmental period of transition in difficulties, such as anxiety or depression. This
which youth strive for more independence from is an important finding, as it suggests that
parents (Steinberg, 2001). It is possible that parent–adolescent conflict has far-reaching
adolescents’ desires for more independence do effects on adolescent well-being and might be
not align with the power structures within some an important risk factor and correlate for comor-
families in which parents maintain power and bid problems among adolescents, such that
influence while youth attempt to transition to a youth experience co-occurring maladjustment
more mutual, egalitarian relationship. Bound- problems (e.g., internalizing and externaliz-
aries in these families may be more relatively ing symptoms). Co-occurrence is common in
closed than open and may be less malleable adolescent psychopathology and has important
to transformation as youth mature. Among implications for the severity and levels of youth
parent–adolescent relationships characterized dysfunction and for the needed complexity of
by higher power asymmetry, in the context of treatments (Graber & Sontag, 2009).
parent–adolescent disagreement youth may feel
that their difference in opinion is not acknowl- Effect-Level Moderators. We also examined
edged or valued, which may contribute to several effect- and study-level moderators of the
higher youth maladjustment. When youth begin association between parent–adolescent conflict
to move out of parents’ homes and become and youth maladjustment. The effect size did
financially independent, the hierarchy of the not vary across study-level moderators of publi-
parent–adolescent relationships might be mod- cation year, youth age, race, or family structure
ified (Hock, Eberly, Bartle-Haring, Ellwanger, and/or composition.
& Widaman, 2001) and the association between For youth gender, moderator analyses
parent–child disagreement and maladjustment revealed that the effect size was significantly
among young adults might change. Examining stronger among samples of boys and girls than
the levels of hierarchy in parent–adolescent samples of only boys or only girls. Moreover,
relationships as a moderator of associations the effect size was significantly stronger among
and the relative contribution of parent–child samples of only girls than samples of only boys.
disagreement and hostility to outcomes among Research has found that girls tend to engage
young adults could be important directions for in more harmony-maintenance activities in
future research. relationships (Davies & Lindsay, 2004), such
Results also indicated that the strength of the that they are more likely than boys to try to
association between parent–adolescent conflict appease others through repression of their own
Parent–Adolescent Conflict 107

thoughts and feelings (Rosenfield, Vertefuille, or adjust to pivotal, transformational char-

& McAlpine, 2000). In contrast, boys may acteristics during adolescence (Steinberg &
be more assertive than girls in relationships Silk, 2002). For example, parent–adolescent
(Rosenfield et al., 2000). The ways boys navi- conflict might signal the need for families to
gate relationships and conflict may buffer some adapt and change, such as adjusting strong
of the effects that parent–adolescent conflict has hierarchies in parent–adolescent relationships,
on their well-being in comparison to girls, such to accommodate adolescents’ increasing needs
that conflict with parents may have a stronger for independence and egalitarianism. In line
impact on girls’ than boys’ attachment bonds with family stress theory (McCubbin & Pat-
with parents and internal working models of terson, 1983), if families struggle to adjust
relationships. Research has found that, with and adapt necessary changes in response to
insecure parent–adolescent attachment rela- parent–adolescent conflict and the develop-
tionships, maternal negativity is associated ment needs of adolescents, the effects of
with youth internalizing behaviors among girls parent–adolescent conflict on youth malad-
but not boys (Milan, Zona, & Snow, 2013). justment might become stronger over time.
Other studies have found that the association Results also indicated that the effect size was
between insecure attachment and hostile feel- stronger for effects of lower methodological
ings declines less over time among girls than quality. This is consistent with research that has
boys (Kobak, Zajac, & Smith, 2009). Boys’ suggested that less methodologically rigorous
tendency to engage in more assertiveness in studies likely overestimate effect sizes (Jüni
relationships might increase parents’ awareness et al., 2001).
of boys’ needs and opinions and contribute to
stronger pushes in developmentally necessary Limitations
changes in the power hierarchies and boundaries
Despite the large number of effects, this study
of parent–adolescent relationships.
was limited by an insufficient number of effects
In contrast to the findings for youth gen-
to conduct more specific analyses. First, we were
der, parent gender did not moderate the asso-
unable to examine specialized effects between
ciation between parent–adolescent conflict and
parent–adolescent conflict and youth malad-
youth maladjustment. A major criticism of the
justment. As such, we were not able to rule out
literature on parent–adolescent conflict is that whether there was a difference in the strength
a minimal number of studies have examined of associations when examining specific dimen-
father–adolescent relationships. Findings from sions of parent–adolescent conflict and specific
this study add to the literature by suggesting dimensions of youth maladjustment. Second,
that the association between parent–adolescent we were not able to test additional three-way
conflict and youth maladjustment might function interactions. It is unclear whether the strength
very similarly across mothers and fathers. A suf- of the association between parent–adolescent
ficient number of effect estimates of fathers were conflict and youth maladjustment varies as a
available for us to be confident in these find- function of the interaction of conflict dimen-
ings. Moreover, these results suggest that any sions, maladjustment dimensions, and specific
gender differences in the association between moderators, such as youth age or the race of the
parent–adolescent conflict and youth maladjust- sample. Moreover, we were not able to examine
ment might be influenced by youths’ rather than associations as a function of parent–adolescent
parents’ characteristics. Methodologically, this gender dyads. Significant differences were
study provides support for this idea, given that found regarding youth gender; however, is it
our results also indicate that the association is unclear how youth and parent genders, in
between parent–adolescent conflict and youth addition to conflict dimensions, may interact to
maladjustment is stronger for youth than parent predict youth maladjustment.
reports of conflict.
The association between parent–adolescent
conflict and youth maladjustment also was Conclusions
stronger for longitudinal effects than for Parent–adolescent conflict is an important
cross-sectional effects. This finding is impor- construct for family, developmental, and
tant, as it suggests that families must adapt clinical studies. Our goal was to clarify the
108 Journal of Family Theory & Review

existing literature on the association between *

Bahrassa, N. F., Juan, M. D., & Lee, R. M. (2013).
parent–adolescent conflict and youth mal- Hmong American sons and daughters: Exploring
adjustment by synthesizing the literature mechanisms of parent–child acculturation con-
and examining the contribution of specific flicts. Asian American Journal of Psychology, 4,
dimensions of conflict to youth maladjust- 100–108. doi: 10.1037/a0028451
*
ment. Findings from this study suggest that Bámaca-Colbert, M. Y., Umaña-Taylor, A. J., &
Gayles, J. G. (2012). A developmental-contextual
parent–adolescent conflict is an important risk model of depressive symptoms in Mexican-origin
factor for youth maladjustment, but associations female adolescents. Developmental Psychology,
do not vary as a function of conflict dimen- 48, 406–421. doi: 10.1037/a0025666
sions. These results are important in terms of Bandura, A. (1973). Aggression: A social learning
future theoretical construction and integration. analysis. Oxford, UK: Prentice Hall.
*
Future research should consider more closely Barber, B. L., & Lyons, J. M. (1994). Family pro-
why and how parent–adolescent disagreement cesses and adolescent adjustment in intact and
can theoretically facilitate both functional, and remarried families. Journal of Youth and Adoles-
seemingly positive, transformational processes cence, 23, 421–436. doi: 10.1007/BF01538037
*
during adolescence in which youth individuate Barrera, M. J., & Stice, E. (1998). Parent–adolescent
conflict in the context of parental support: Fam-
from their parents, as suggested by theory (Blos,
ilies with alcoholic and nonalcoholic fathers.
1979; Grotevant & Cooper, 1985), and the more Journal of Family Psychology, 12, 195–208. doi:
negative indicators of youth maladjustment 10.1037/0893-3200.12.2.195
observed in the findings of this study. Integration Blos, P. (1979). The adolescent passage. New York,
of theories that focus on the topics and mean- NY: International Universities Press.
ings of conflict for parents and youth (Smetana, Bowen, M. (1978). Family therapy in clinical prac-
1996) with theories that discuss the power tice. New York, NY: Aronson.
dynamics in parent–adolescent relationships Bowlby, J. (1973). Attachment and loss: Vol. 2. Sep-
(Minuchin, 1974) may offer promising frame- aration: anxiety and anger. New York, NY: Basic
works for future research. Moreover, this study Books.
*
Bradford, K., Vaughn, L., & Barber, B. K. (2008).
provides evidence that parent–adolescent con-
When there is conflict: Interparental conflict,
flict has implications for youth maladjustment parent–child conflict, and youth problem behav-
over time. How families manage transitions iors. Journal of Family Issues, 29, 780–805. doi:
and adjust parent–adolescent relationships and 10.1177/0192513X07308043
boundaries in response to parent–adolescent *
Briere, F. N., Archambault, K., & Janosz, M. (2013).
conflict may have important implications for Reciprocal prospective associations between
later youth maladjustment. Future research depressive symptoms and perceived relationship
should extend these findings and examine with parents in early adolescence. Canadian
whether parent–adolescent conflict dimensions Journal of Psychiatry, 58, 169–176.
may have specialized effects on particular youth Bronstein, P. (1984). Differences in mothers’
outcomes over time. and fathers’ behaviors toward children: A
cross-cultural comparison. Developmental Psy-
chology, 20, 995–1003. 10.1037/0012-1649.20.
References 6.995
Buehler, C. (2006). Parents and peers in relation
References marked with an asterisk indicate studies to early adolescent problem behavior. Journal
included in the meta-analysis. of Marriage and Family, 68, 109–124. doi:
Achenbach, T. M., & Edelbrock, C. S. (1978). 10.1111/j.1741-3737.2006.00237.x
The classification of child psychopathology: Buehler, C., Anthony, C., Krishnakumar, A., Stone,
A review and analysis of empirical efforts. Psy- G., Gerard, J., & Pemberton, S. (1997). Inter-
chological Bulletin, 85, 1275–1301. doi: 10.1037/ parental conflict and youth problem behaviors: A
0033-2909.85.6.1275 meta-analysis. Journal of Child and Family Stud-
* ies, 6, 233–247.
Acock, A. C., & Demo, D. H. (1999). Dimensions
*
of family conflict and their influence on child and Buehler, C., & Gerard, J. M. (2002). Marital con-
adolescent adjustment. Sociological Inquiry, 69, flict, ineffective parenting, and children’s and ado-
641–658. lescents’ maladjustment. Journal of Marriage and
*
Adams, R. E., & Laursen, B. (2007). The corre- Family, 64, 78–92.
lates of conflict: Disagreement is not necessar- Buehler, C., Krishnakumar, A., Anthony, C.,
ily detrimental. Journal of Family Psychology, 21, Tittsworth, S., & Stone, G. (1994). Hostile
445–458. doi: 10.1037/0893-3200.21.3.445 interparental conflict and youth maladjustment.
Parent–Adolescent Conflict 109

Family Relations: An Interdisciplinary Journal *

Cunningham, K. S. (2008). Moderators of the
of Applied Family Studies, 43, 409–416. doi: relationship of family conflict and depressive
10.2307/585372 symptomatology in adolescents. Dissertation
*
Burk, W. J. (2005). Concurrent and prospective Abstracts International, 68, 7707.
associations between parent–adolescent conflict Davies, P. T., & Lindsay, L. L. (2004). Interparental
and adolescent adjustment: Person-oriented and conflict and adolescent adjustment: Why does
variable-oriented analyses. Dissertation Abstracts gender moderate early adolescent vulnerability?
International, 66, 3440. Journal of Family Psychology, 18, 160–170. doi:
*
Burt, S. A. (2004). The nature of the association 10.1037/0893-3200.18.1.160
between parent–child conflict and childhood exter- Demo, D. H., & Fine, M. A. (2010). Beyond the
nalizing pathology. Dissertation Abstracts Interna- average divorce. Thousand Oaks, CA: Sage.
tional, 65, 2088. Dixon, S. V., Graber, J. A., & Brooks-Gunn, J. (2008).
Campione-Barr, N., & Smetana, J. G. (2010). “Who Roles of respect for parental authority and par-
said you could wear my sweater?” Adolescent enting practices in parent–child conflict among
siblings’ conflicts and associations with relation- African American, Latino and European American
ship quality. Child Development, 81, 464–471. doi: families. Journal of Family Psychology, 22, 1–10.
10.1111/j.1467-8624.2009.01407.x doi: 10.1037/0893-3200.22.1.1
* *
Caples, H. S., & Barrera, M., Jr. (2006). Con- Duncan, S. C., Duncan, T. E., Biglan, A., & Ary,
flict, support and coping as mediators of the rela- D. (1998). Contributions of the social context
tion between degrading parenting and adolescent to the development of adolescent substance use:
adjustment. Journal of Youth and Adolescence, 35, A multivariate latent growth modeling approach.
599–611. Drug and Alcohol Dependence, 50, 57–71. doi:
Cicchetti, D., & Rogosch, F. A. (2002). A 10.1016/S0376-8716(98)00006-4
developmental psychopathology perspec- *
Ehrlich, K. B., Dykas, M. J., & Cassidy, J. (2012).
tive on adolescence. Journal of Consulting Tipping points in adolescent adjustment: Predict-
and Clinical Psychology, 70, 6–20. doi: ing social functioning from adolescents’ conflict
10.1037/0022-006X.70.1.6 with parents and friends. Journal of Family Psy-
*
Choi, Y., He, M., & Harachi, T. W. (2008). Intergen- chology, 26, 776–783. doi: 10.1037/a0029868
erational cultural dissonance, parent–child conflict *
El-Sheikh, M., & Elmore-Staton, L. (2004). The
and bonding, and youth problem behaviors among link between marital conflict and child adjustment:
Vietnamese and Cambodian immigrant families. Parent–child conflict and perceived attachments
Journal of Youth and Adolescence, 37, 85–96. doi: as mediators, potentiators, and mitigators of risk.
10.1007/s10964-007-9217-z Development and Psychopathology, 16, 631–648.
*
Chung, G. H., Flook, L., & Fuligni, A. J. (2009). doi: 10.1017/S0954579404004705
Daily family conflict and emotional distress Faber, A. J. (2002). The role of hierarchy in parental
among adolescents from Latin American, Asian, nurturance. American Journal of Family Therapy,
and European backgrounds. Developmental Psy- 30, 73–84. doi: 10.1080/019261802753455660
chology, 45, 1406–1415. doi: 10.1037/a0014163 *
Flannery, D. J. (1991). The impact of puberty on
*
Constantine, M. G. (2006). Perceived family conflict, parent–adolescent relations: An observational
parental attachment, and depression in African study of the relationship between affect and
American female adolescents. Cultural Diversity engagement in interactions, parent–adolescent
and Ethnic Minority Psychology, 12, 697–709. doi: conflict, and adolescent problem behavior. Disser-
10.1037/1099-9809.12.4.697 tation Abstracts International, 52, 1688.
* *
Costigan, C. L., Cauce, A. M., & Etchison, K. (2007). Ford, A. B. (1997). Post divorce parental conflict, the
Changes in African American mother-daughter coparental relationship, and the parent–adolescent
relationships during adolescence: Conflict, auton- relationship in predicting adolescent adjustment to
omy, and warmth. In B. R. Leadbeater & N. Way divorce. Dissertation Abstracts International, 57,
(Eds.), Urban girls revisited: Building strengths 6570.
(pp. 177–201). New York, NY: New York Univer- Fuligni, A. J. (1998). Authority, autonomy, and
sity Press. parent–adolescent conflict and cohesion: A
*
Crean, H. F. (2008). Conflict in the Latino study of adolescents from Mexican, Chi-
parent–youth dyad: The role of emotional nese, Filipino, and European backgrounds.
support from the opposite parent. Journal Developmental Psychology, 34, 782–792. doi:
of Family Psychology, 22, 484–493. doi: 10.1037/0012-1649.34.4.782
10.1037/0893-3200.22.3.484 García Coll, C. G., Meyer, E. C., & Brillon L. (1995).
Cummings, E., Davies, P. T., & Campbell, S. B. Ethnic and minority parenting. In M. H. Bornstein
(2000). Developmental psychopathology and fam- (Ed.), Handbook of parenting: Vol. 2. Biology and
ily process: Theory, research, and clinical implica- ecology of parenting (pp. 189–209). Mahwah, NJ:
tions. New York, NY: Guilford Press. Erlbaum.
110 Journal of Family Theory & Review

Gavazzi, S. M., & Sabatelli, R. M. (1990). the development of personality disorders. Devel-
Family system dynamics, the individuation opment and Psychopathology, 21, 839–851. doi:
process, and psychosocial development. Jour- 10.1017/S0954579409000455
nal of Adolescent Research, 5, 500–519. doi: Krishnakumar, A., & Buehler, C. (2000). Inter-
10.1177/074355489054008 parental conflict and parenting behaviors: A meta
Goldhaber, D. E. (2000). Theories of human devel- analytic review. Family Relations, 49, 25–44. doi:
opment: Integrative perspectives. Mountain View, 10.1111/j.1741-3729.2000.00025.x
*
CA: Mayfield. Krishnan, U. D. (2005). Parent–adolescent conflict
*
Gonzales, N. A. (1993). Parent–adolescent relations and adolescent functioning in a collectivist, ethni-
and adolescent adjustment in African-American cally heterogenous culture: Malaysia. Dissertation
families. Dissertation Abstracts International, 53, Abstracts International, 65, 4493.
*
4975. Kuhlberg, J. A., Peña, J. B., & Zayas, L. H. (2010).
Gonzales, N. A., Cauce, A., & Mason, C. A. (1996). Familism, parent–adolescent conflict, self-esteem,
Interobserver agreement in the assessment of internalizing behaviors and suicide attempts
parental behavior and parent–adolescent con- among adolescent Latinas. Child Psychiatry
flict: African American mothers, daughters, and and Human Development, 41, 425–440. doi:
independent observers. Child Development, 67, 10.1007/s10578-010-0179-0
1483–1498. doi: 10.2307/1131713 Kuppens, S., Laurent, L., Heyvaert, M., & Onghena,
Graber, J. A., & Sontag, L. M. (2009). Internalizing P. (2013). Associations between parental psy-
problems during adolescence. In R. M. Lerner chological control and relational aggression in
& L. Steinberg (Eds.), Handbook of adolescent children and adolescents: A multilevel and sequen-
psychology: Vol. 1. Individual bases of adolescent tial meta-analysis. Developmental Psychology, 49,
development (3rd ed., pp. 642–682). Hoboken, NJ: 1697–1712. doi: 10.1037/a0030740
Wiley. Laursen, B., Coy, K. C., & Collins, W. (1998). Recon-
*
Greenley, R. N., Taylor, H. G., Drotar, D., & Minich, sidering changes in parent–child conflict across
N. M. (2007). Longitudinal relationships between adolescence: A meta-analysis. Child Development,
early adolescent family functioning and youth 69, 817–832. doi: 10.2307/1132206
*
adjustment: An examination of the moderating role Lewandowski, A. S., & Palermo, T. M. (2009).
of very low birth weight. Journal of Pediatric Psy- Parent–teen interactions as predictors of depres-
chology, 32, 453–462. doi: 10.1093/jpepsy/jsl027 sive symptoms in adolescents with headache. Jour-
Grotevant, H. D., & Cooper, C. R. (1985). Patterns of nal of Clinical Psychology in Medical Settings, 16,
interaction in family relationships and the develop- 331–338. doi: 10.1007/s10880-009-9173-8
ment of identity exploration in adolescence. Child Littell, R. C., Milliken, G. A., Stroup, W. W., Wolfin-
Development, 56, 415–428. doi: 10.2307/1129730 ger, R. D., & Schabenberger, O. (2006). SAS for
*
Hakvoort, E. M., Bos, H. W., van Balen, F., & Her- mixed models (2nd ed.). Cary, NC: SAS Institute.
*
manns, J. A. (2010). Family relationships and the Maggs, J. L., & Galambos, N. L. (1993). Alter-
psychosocial adjustment of school-aged children native structural models for understanding ado-
in intact families. Journal of Genetic Psychology: lescent problem behavior in two-earner families.
Research and Theory on Human Development, Journal of Early Adolescence, 13, 79–101. doi:
171, 182–201. doi: 10.1080/00221321003657445 10.1177/0272431693013001005
*
Hock, E., Eberly, M., Bartle-Haring, S., Ellwanger, Manne, S., & Miller, D. (1998). Social support,
P., & Widaman, K. F. (2001). Separation anxiety social conflict, and adjustment among adolescents
in parents of adolescents: Theoretical significance with cancer. Journal of Pediatric Psychology, 23,
and scale development. Child Development, 72, 121–130. doi: 10.1093/jpepsy/23.2.121
284–298. doi: 10.1111/1467-8624.00279 Marsh, H. W., Bornmann, L., Mutz, R., Daniel, H.,
*
Juang, L. P., Syed, M., & Cookston, J. T. & O’Mara, A. (2009). Gender effects in the peer
(2012). Acculturation-based and everyday reviews of grant proposals: A comprehensive
parent–adolescent conflict among Chinese Amer- meta-analysis comparing traditional and multi-
ican adolescents: Longitudinal trajectories and level approaches. Review of Educational Research,
implications for mental health. Journal of Family 79, 1290–1326. doi: 10.3102/0034654309334143
Psychology, 26, 916–926. doi: 10.1037/a0030057 McCubbin, H. I., & Patterson, J. M. (1983). Family
Jüni, P., Altman, D. G., & Egger, M. (2001). System- transitions: Adaptation to stress. In H. I. McCub-
atic reviews in health care: Assessing the quality of bin & C. R. Figley (Eds.), Stress and the family:
controlled clinical trials. British Medical Journal, Coping with normative transitions (pp. 5–25). New
323, 42–46. doi: 10.1136/bmj.323.7303.42 York, NY: Brunner/Mazel.
*
Kobak, R., Zajac, K., & Smith, C. (2009). McKinney, C., & Renk, K. (2011). A multivariate
Adolescent attachment and trajectories of model of parent–adolescent relationship vari-
hostile-impulsive behavior: Implications for ables in early adolescence. Child Psychiatry
Parent–Adolescent Conflict 111

and Human Development, 42, 442–462. doi: effects. Journal of School Violence, 7, 3–25. doi:
10.1007/s10578-011-0228-3 10.1300/J202v07n01_02
Milan, S., Zona, K., & Snow, S. (2013). Pathways Smetana, J. G. (1996). Adolescent–parent con-
to adolescent internalizing: Early attachment inse- flict: Implications for adaptive and maladaptive
curity as a lasting source of vulnerability. Journal development. In D. Cicchetti & S. L. Toth (Eds.),
of Clinical Child and Adolescent Psychology, 42, Adolescence: Opportunities and challenges (pp.
371–383. doi: 10.1080/15374416.2012.736357 1–46). Rochester, NY: University of Rochester
Minuchin, S. (1974). Families and family therapy. Press.
Boston, MA: Harvard University Press. Smetana, J. G. (2005). Adolescent–parent conflict:
*
Pasch, L. A., Deardorff, J., Tschann, J. M., Flores, Resistance and subversion as developmental pro-
E., Penilla, C., & Pantoja, P. (2006). Acculturation, cess. In L. Nucci (Ed.), Conflict, contradiction,
parent–adolescent conflict, and adolescent adjust- and contrarian elements in moral development and
ment in Mexican- American families. Family Pro- education (pp. 69–91). Mahwah, NJ: Erlbaum.
cess, 45, 75–86. doi: 10.1111/j.1545-5300.2006. Smetana, J. G., Abernethy, A., & Harris, A. (2000).
00081.x Adolescent–parent interactions in middle-class
Peterson, R. A., & Brown, S. P. (2005). On the African American families: Longitudinal change
use of beta coefficients in meta-analysis. Jour- and contextual variations. Journal of Family Psy-
nal of Applied Psychology, 90, 175–181. doi: chology, 14, 458–474. doi: 10.1037/0893-3200.
10.1037/0021-9010.90.1.175 14.3.458
Proulx, C. M., Helms, H. M., & Buehler, C. (2007). Smetana, J. G., Campione-Barr, N., & Metzger,
Marital quality and personal well-being: A A. (2006). Adolescent development in inter-
meta-analysis. Journal of Marriage and Family, personal and societal contexts. Annual Review
69, 576–593. doi: 10.1111/j.1741-3737.2007. of Psychology, 57, 255–284. doi: 10.1146/
00393.x [Link].57.102904.190124
Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchi- Smetana J. G., Crean H. F., & Daddis C. (2002).
cal linear models: Applications and data analysis Family processes and problem behaviors in
middle-class African American adolescents. Jour-
methods. Thousand Oaks, CA: Sage.
* nal of Research on Adolescence, 12, 275–304. doi:
Richmond, M. K., & Stocker, C. M. (2006).
10.1111/1532-7795.00034
Associations between family cohesion and
Smetana, J. G., Yau, J., Restrepo, A., & Braeges, J. L.
adolescent siblings’ externalizing behavior. Jour-
(1991). Adolescent–parent conflict in married and
nal of Family Psychology, 20, 663–669. doi:
divorced families. Developmental Psychology, 27,
10.1037/0893-3200.20.4.663
1000–1010. doi: 10.1037/0012-1649.27.6.1000
Rosenfield, S., Vertefuille, J., & McAlpine, D. *
Smith, T. L. (2001). Linking supportive-involved
D. (2000). Gender stratification and mental parenting, parent–child conflict, peer orienta-
health: An exploration of dimensions of the self. tion, peer relationship quality, and academic
Social Psychology Quarterly, 63, 208–223. doi: self-efficacy to the academic achievement of rural
10.2307/2695869 African American youth. Dissertation Abstracts
Rosenthal, R. (1991). Meta-analytic procedures for International Section A, 61, 3894.
social research. Thousand Oaks, CA: Sage. Sprey, J. (1969). The family as a system in conflict.
*
Rueter, M. A., Scaramella, L., Wallace, L. E., & Con- Journal of Marriage and the Family, 31, 699–706.
ger, R. D. (1999). First onset of depressive or anx- *
Steeger, C. M., & Gondoli, D. M. (2013).
iety disorders predicted by the longitudinal course Mother–adolescent conflict as a mediator between
of internalizing symptoms and parent–adolescent adolescent problem behaviors and maternal psy-
disagreements. Archives of General Psychiatry, chological control. Developmental Psychology,
56, 726–732. doi: 10.1001/archpsyc.56.8.726 49, 804–814. doi: 10.1037/a0028599
*
Shek, D. L. (1998). A longitudinal study of the Steinberg, L. (1990). Autonomy, conflict, and har-
relations between parent–adolescent conflict and mony in the family relationship. In S. Feldman &
adolescent psychological well-being. Journal G. R. Elliott (Eds.), At the threshold: The devel-
of Genetic Psychology: Research and The- oping adolescent (pp. 255–276). Cambridge, MA:
ory on Human Development, 159, 53–67. doi: Harvard University Press.
10.1080/00221329809596134 Steinberg, L. (2001). We know some things:
*
Shek, D. L., & Ma, H. K. (2001). Parent–adolescent Parent–adolescent relationships in retrospect and
conflict and adolescent antisocial and prosocial prospect. Journal of Research on Adolescence, 11,
behavior: A longitudinal study in a Chinese con- 1–19. doi: 10.1111/1532-7795.00001
text. Adolescence, 36, 545–555. Steinberg, L., & Silk, J. S. (2002). Parenting adoles-
*
Simons-Morton, B., Chen, R., Hand, L., & Haynie, cents. In M. H. Bornstein (Ed.), Handbook of par-
D. L. (2008). Parenting behavior and adolescent enting: Vol. 1. Children and parenting (2nd ed., pp.
conduct problems: Reciprocal and mediational 103–133). Mahwah, NJ: Erlbaum.
112 Journal of Family Theory & Review

*
Stevens, G. M., Vollebergh, W. M., Pels, T. M., & Pedersen, S. L. (2012). Substance use and
& Crijnen, A. M. (2007). Problem behav- delinquency among adolescents with childhood
ior and acculturation in Moroccan immigrant ADHD: The protective role of parenting. Psy-
adolescents in the Netherlands: Effects of chology of Addictive Behaviors, 26, 585–598. doi:
gender and parent–child conflict. Journal of 10.1037/a0026818
*
Cross-Cultural Psychology, 38, 310–317. doi: Wang, M., Brinkworth, M., & Eccles, J. (2013).
10.1177/0022022107300277 Moderating effects of teacher–student relationship
*
Stewart, S., Lam, T. H., Betson, C., & Chung, S. in adolescent trajectories of emotional and behav-
F. (1999). Suicide ideation and its relationship to ioral adjustment. Developmental Psychology, 49,
depressed mood in a community sample of adoles- 690–705. doi: 10.1037/a0027916
*
cents in Hong Kong. Suicide and Life-Threatening Yeh, K. (2011). Mediating effects of nega-
Behavior, 29, 227–240. tive emotions in parent–child conflict on
*
Summers, P., Forehand, R., Armistead, L., & Tannen- adolescent problem behavior. Asian Jour-
baum, L. (1998). Parental divorce during early ado- nal of Social Psychology, 14, 236–245. doi:
lescence in Caucasian families: The role of family 10.1111/j.1467-839X.2011.01350
*
process variables in predicting the long-term con- Yin, X., Li, Z., & Su, L. (2012). Fathers’ parent-
sequences for early adult psychosocial adjustment. ing and children’s adjustment: The mediating
Journal of Consulting and Clinical Psychology, role of father-child conflict. Social Behav-
66, 327–336. doi: 10.1037/0022-006X.66.2.327 ior and Personality, 40, 1401–1408. doi:
Tesser, A., Forehand, R., Brody, G., & Long, N. 10.2224/sbp.2012.40.8.1401
(1989). Conflict: The role of calm and angry Yuan, K., & Bentler, P. M. (2002). On normal theory
parent–child discussion in adolescent adjustment. based inference for multilevel models with distri-
Journal of Social and Clinical Psychology, 8, butional violations. Psychometrika, 67, 539–562.
317–330. doi: 10.1521/jscp.1989.8.3.317 doi: 10.1007/BF02295130
*
Torgerson, C. J. (2006). Publication bias: The Zeiders, K. H., Roosa, M. W., & Tein, J. (2011).
Achilles’ heel of systematic reviews? British Family structure and family processes in
Journal of Educational Studies, 54, 89–102. doi: Mexican-American families. Family Process, 50,
10.1111/j.1467-8527.2006.00332.x 77–91. doi: 10.1111/j.1545-5300.2010.01347.x
*
Updegraff, K. A., Umaña-Taylor, A. J., Perez-Brena,
N. J., & Pflieger, J. (2012). Mother–daughter con-
flict and adjustment in Mexican-origin families: Supporting Information
Exploring the role of family and sociocultural Additional supporting information may be
context. In L. P. Juang & A. J. Umaña-Taylor found in the online version of this article at
(Eds.), Family conflict among Chinese- and
[Link]:
Mexican-origin adolescents and their parents
in the U.S. (pp. 59–81). San Francisco, CA: Table A1. Summary of Studies for the Rela-
Jossey-Bass. tionship between Parent-Adolescent Conflict and
Van den Noortgate, W., & Onghena, P. (2003). Multi- Youth Maladjustment
level meta-analysis: A comparison with traditional Table A2. Sensitivity Analyses Using the
meta-analytical procedures. Educational and Sandwich Estimator
Psychological Measurement, 63, 765–790. doi: Figure A1. Funnel plot of Fisher’s Z trans-
10.1177/0013164402251027 formed effect sizes.
*
Walther, C. P., Cheong, J., Molina, B. G., Pel-
ham, W. J., Wymbs, B. T., Belendiuk, K. A.,