CHERYL BUEHLER AND JEAN M.

GERARD
University of Tennessee

Marital Conflict, Ineffective Parenting, and Children’s

and Adolescents’ Maladjustment

Data from the 1988 National Survey on Families cioemotional development directly by shaping
and Households were analyzed to examine the as- children’s cognitions and perceptions (Davies &
sociations among marital conflict, ineffective par- Cummings, 1998; Grych & Fincham, 1990), their
enting, and children’s and adolescents’ malad- ability to regulate their emotions (Gottman &
justment. Parents’ use of harsh discipline and low Katz, 1989), and their particular coping responses
parental involvement helped explain the connec- (Jenkins, Smith, & Graham, 1989; O’Brien, Mar-
tion between marital conflict and children’s mal- golin, & John, 1995). Scholars also suggest that
adjustment in children aged 2 through 11. Parent- marital conflict affects children’s maladjustment
child conflict was measured only in families with indirectly by altering parenting practices and the
a target teenager and also was a significant me- quality of parent-child relations (Erel & Burman,
diator. Although ineffective parenting explained 1995; Fauber, Forehand, Thomas, & Wierson,
part of the association between marital conflict 1990). The purpose of this study is to examine
and children’s maladjustment, independent effects this proposition that marital conflict is associated
of marital conflict remained in families with target with children’s maladjustment, in part through its
children aged 2 through 11 (but not for families association with ineffective parenting.
with a teenager). With a few exceptions, this pat-
tern of findings was consistent for mothers’ and LITERATURE REVIEW
fathers’ reports, for daughters and sons, for fam-
ilies with various ethnic backgrounds, and for The idea that marital conflict might shape ineffec-
families living in and out of poverty. tive parenting (and ultimately children’s malad-
justment) often is referred to as ‘‘spillover.’’ Spill-
over is defined as the direct transfer of mood,
Exposure to marital conflict is associated with a affect, or behavior from one setting to another
range of indicators of children’s maladjustment (Engfer, 1988; Erel & Burman, 1995; Repetti,
(c.f., Brody, Stoneman, & Flor, 1995; Grych, Seid, 1987). Our interest in examining the potential
& Fincham, 1992; Harold, Fincham, Osborne, & spillover from conflictual marital interactions to
Conger, 1997; Katz & Gottman, 1993). Scholars ineffective parenting comes from two sources.
suggest that marital conflict affects children’s so- The first is Fincham’s (1994) belief that it is time
to move onto ‘‘second-generation’’ research in the
Department of Child and Family Studies, University of area of marital conflict. One of the aims of sec-
Tennessee, 1215 West Cumberland Avenue, Room 115, ond-generation research is to explain how and
Knoxville, TN 37996-1900 ([email protected]).
why marital conflict and children’s maladjustment
Key Words: adolescent maladjustment, children’s malad- are associated by examining it in conjunction with
justment, marital conflict, parenting, spillover. other risk factors. Within the context of marital

78 Journal of Marriage and Family 64 (February 2002): 78–92

Marital Conflict 79

conflict, ineffective parenting is one of these sa- 1986; Stone, Buehler, & Barber, in press; Vande-
lient risk factors (Fauber & Long, 1991). A sec- water & Lansford, 1998). Three of these studies
ond inspiration for this theoretical proposition presented data from two samples, so the total
comes from Belsky’s (1984) model of the deter- number of examinations was 15; 13 of 15 found
minants of parenting. In this model, he suggested evidence of a positive association between marital
that parenting is a central, proximal socialization conflict and children’s maladjustment.
influence in children’s development and that child Seven of the studies provided support for the
and parental characteristics shape parenting. One proposition that ineffective parenting at least par-
of the salient predictors of parenting is the quality tially mediates the association between marital
of the marital relationship. Theoretically, support conflict and children’s maladjustment (Acock &
from one’s marital partner facilitates effective par- Demo, 1999; Harold & Conger, 1997; Harold et
enting, whereas marital conflict disrupts it. al., 1997; Harrist & Ainslie, 1998; Mann &
Several recent reviews of marital conflict re- MacKenzie, 1996; Osborne & Fincham, 1996;
search discuss the idea of spillover and the poten- Vandewater & Lansford, 1998). Depending on the
tial mediating role of ineffective parenting particular analysis, parenting was both a partial
(Crockenberg & Covey, 1991; Davies & Cum- and full mediator. The fact that some analyses
mings, 1994; Emery & Tuer, 1993; Erel & Bur- suggested full effects and others only partial ef-
man, 1995; Grych & Fincham, 1990; Sanders, fects did not seem to be a function of children’s
Nicholson, & Floyd, 1997; Wilson & Gottman, ages (although only a few studies examined chil-
1995). Most recently, Krishnakumar and Buehler dren under the age of 10), type of problem be-
(2000) documented an association between mari- havior (internalizing or externalizing), nor specific
tal conflict and ineffective parenting in a meta- indicator of parenting (although most studies just
analysis of 138 effect coefficients from 39 studies. examined the affective nature of parent-child re-
They reported that the average weighted effect lations, such as rejection, hostility, or conflict).
size (d) was .62 (r 5 .30). Associations were In six of these samples at least some of the
strongest when correlating marital conflict and pa- analyses refuted the idea that parenting mediates
rental harsh punishment or lack of acceptance of the association between marital conflict and chil-
the child. Using daily checklists to record marital dren’s maladjustment (Harold & Conger, 1997;
conflict and parent-child conflict, Almeida, Weth- Miller et al., 1993; Peterson & Zill, 1986; Stone
ington, and Chandler (1999) corroborated the et al., in press). Thus, mediating effects might ex-
findings from the meta-analysis by reporting that ist, but they are not a foregone conclusion based
both mothers and fathers are about 50% more on the empirical literature.
likely to have tense interaction with their children Three of the studies found support for indirect
on the day after there had been some marital ten- effects rather than mediating. Marital conflict, in-
sion. effective parenting, and children’s maladjustment
The idea that marital conflict shapes ineffective were linked, but either the direct effect between
parenting and ultimately children’s maladjustment marital conflict and children’s maladjustment did
is fairly widely accepted in the marital conflict not exist (Gonzalez et al., 2000; Harold & Conger,
literature. In their review, Crockenberg and Covey 1997) or the direct effect did not diminish when
(1991) stated that ‘‘there is little question at this parenting was considered (Stone et al., in press
point that marital conflict affects child external- [two samples]). A few other studies that did not
izing behavior indirectly, through its link with pa- establish empirically an association between mar-
rental behavior. Whether it has a direct effect on ital conflict and children’s maladjustment before
the child as well is less certain’’ (p. 250). Given considering parenting also have documented in-
the strength of this statement, we were surprised direct effects (Anderson, Lindner, & Bennion,
to find only 12 published or in press articles or 1992; Fauber et al., 1990).
book chapters that have examined the mediating As a whole, this literature is limited in that (a)
effects of ineffective parenting (Acock & Demo, studies including young children are few; (b) most
1999; Cowan, Cowan, Heming, & Miller, 1991; studies use regionally based, convenience samples
Gonzales, Pitts, Hill, & Roosa, 2000; Harold & (although some demonstrate representativeness at
Conger, 1997; Harold et al., 1997; Harrist & Ain- the local level), and (c) most studies focus on only
slie, 1998; Mann & MacKenzie, 1996; Miller, one aspect of ineffective parenting. This study
Cowan, Cowan, Hetherington, & Clingempeel, strengthens the relatively small extant literature on
1993; Osborne & Fincham, 1996; Peterson & Zill, the indirect role of parenting in four ways: (a) data
80 Journal of Marriage and Family

from a fairly large, nationally representative sam- dent-additive model, and the interactive model. In
ple of families are analyzed, (b) families with chil- the redundancy model, marital conflict and inef-
dren aged 2 through 18 are examined, (c) parent- fective parenting would measure similar phenom-
ing is measured as a multidimensional construct, enon (e.g., the inability to function well in inti-
and (d) competing models to the spillover hypoth- mate relationships). The two would be highly
esis are examined (i.e., the redundant model, the correlated and would not explain unique portions
independent-additive model, the interactive mod- of variance in children’s maladjustment. In the in-
el). dependent-additive model, marital conflict and in-
We hypothesize a hybrid of the spillover and effective parenting would be distinct characteris-
direct effects model in which ineffective parenting tics (although somewhat correlated because both
partially mediates the positive association between are central aspects of family functioning). Each
marital conflict and children’s maladjustment. We would uniquely explain variance in children’s
hypothesize that this model fits for children of maladjustment, and the variance explained by the
varying ages (across ages 2–18) and applies two would be greater than that explained by either
equally well to sons and daughters. Buehler et al. one considered by itself. In the interactive model,
(1997) found no systematic age differences in the the relationship between marital conflict and chil-
68 studies analyzed in their meta-analysis, sup- dren’s maladjustment would be conditionalized by
porting Cummings and Davies (1994) earlier con- the level of ineffective parenting. The pattern
clusion that children of varying ages seem to be most often suggested in this literature is that the
equally vulnerable to the effects of marital con- positive association between marital conflict and
flict. In terms of gender, some scholars believe children’s maladjustment might be smaller when
that boys and girls see and interpret their social- effective parenting is present.
izing environment through a gender filter (for a
review, see Peterson, Bodman, Bush, & Madden-
METHOD
Derdich, 2000) and that the strength of this filter
might increase as youth move into adolescence
Sample
(Linver & Silverberg, 1997). However, the pre-
vailing ideas regarding the nature and effect of The sample for this study was selected from sub-
these filters often contradict. For example, some jects who participated in the 1988 National Sur-
theorists (e.g., Gilligan, 1982) suggest that daugh- vey of Families and Households (NSFH). Data
ters might be more vulnerable to family stressors were collected from a probability sample of
because they are socialized to caretake and safe- 13,017 households in 1987 and 1988. The main
guard family relationships. Theoretically, this sample consists of a random selection of 9,643
caretaking will bring them into the family and households. The over sample consists of a double
marriage during a time when marital tension is sample of African Americans, Puerto Rican
elevated. On the other hand, others (e.g., Reid & Americans, Mexican Americans, single-parent
Crisafulli, 1990) have speculated that sons might families, families with stepchildren, and cohabi-
be more vulnerable because of their sensitivity to tating or recently married couples. The response
and greater difficulty in handling arousal and rate was 73.5% for the main sample and 76.8%
stress. There has been some empirical support for in the over sample (Sweet, Bumpass, & Call,
sons’ vulnerability to disrupted family relation- 1992).
ships (Rothbaum & Weisz, 1994). Thus, different The weighted sample for this study includes
theoretical and empirical positions regarding child 2,541 married parents with a target child aged 2
gender would predict different moderational ef- to 18 living in the household. Stepfamilies and
fects. Because existing research on age and gender single-parent families were not included so that
differences in correlations (rather than means) the data on marital conflict could be interpreted
have produced null or conflicting results, we hy- accurately as information about the marriage to
pothesized similar pathways for sons and daugh- which the child was exposed on a regular basis.
ters (Buehler et al.; Kim, Hetherington, & Reiss, Parents provided all of the information in the
1999; Krishnakumar & Buehler, 2000; Lytton & NSFH. Data were analyzed with provided sample
Romney, 1991). weights to ensure a sample representative of the
Alternative models to the spillover hypothesis married, nonstepfamily U.S. population.
also are examined. Three additional models are Because the parenting and children’s malad-
examined: the redundancy model, the indepen- justment measures varied slightly with the age of
Marital Conflict 81

the target child in the NSFH, the sample was di- cipline and the three involvement items indicated
vided into three groups: families with a target two factors, providing evidence of discriminant
child aged 2 through 4 (unweighted n 5 586; validity of these two aspects of parenting. In ad-
weighted n 5 623), families with a target child dition, the two were not correlated (see Table 1).
aged 5 through 11 (unweighted n 5 815; weighted Construct validity was indicated by significant
n 5 974), and families with a target child aged correlations in the expected direction with paren-
12 through 18 (unweighted n 5 684, weighted n tal depressive affect (harsh discipline r 5 .18, p
5 944). Looking first at the families with a target , .001; involvement r 5 2.15, p , .001).
child aged 2 through 4 and using weighted data, Three aspects of parenting were measured in
54% of the respondents were mothers, 48% of the the sample of families with a target child aged 5
children were daughters, 83% were European through 11: harsh discipline, involvement, and pa-
American, and 7% of the families were poor (i.e., rental presence. Harsh discipline was measured
lived below the poverty line). In the families with using three items (same items as for families with
a target child aged 5 through 11, 51% of the re- preschoolers). Cronbach’s alpha was .93. Involve-
spondents were mothers, 50% of the children were ment was measured using 6 items. These ques-
daughters, 78% were European American, and 7% tions addressed the frequency of shared activities,
of the families were poor. In the families with a projects, and private talks, as well as helping the
target child aged 12 through 18, 51% of the re- child with reading, praising, and hugging. Cron-
spondents were mothers, 48% of the children were bach’s alpha was .70. Parental presence was mea-
daughters, 83% were European American, and 7% sured using five items that addressed how often
of the families were poor. the child was left alone at various times during
the day and night. Items were standardized, a
summed score was calculated, and higher scores
Measures
represented greater parental presence (true for
Marital conflict. Marital conflict was measured us- both older age groups of families). Cronbach’s al-
ing nine items. Six of these asked respondents to pha was .79. A factor analysis that included the
report the frequency of various disagreements dur- three harsh discipline, the six involvement items,
ing the past year. Two of the items measured the and the five parental presence items indicated
frequency of arguing and hitting. One item mea- three factors, providing evidence of discriminant
sured the frequency of calm discussions (reverse validity. The correlations among parenting mea-
coded). Items were standardized because the re- sures were statistically significant but low (see Ta-
sponse formats differed; a mean score was cal- ble 1). The construct validity of the involvement
culated, and higher scores represented higher lev- measure was indicated by a significant negative
els of marital conflict. Cronbach’s alpha was .77. correlation with parental depressive affect (r 5
2.16, p , .001).
Ineffective parenting. Based on the availability of Four aspects of parenting were measured in the
measures, two aspects of ineffective parenting sample of families with a target child aged 12
were measured in the sample of families with a through 18: harsh discipline, parent-adolescent
target child aged 2 through 4: harsh discipline and conflict (parent-child conflict was not measured in
involvement. Harsh discipline was measured us- the NSFH in families with younger children), in-
ing three items. These questions addressed re- volvement, and parental presence. Harsh disci-
spondent’s self-reports of spanking and slapping pline was measured using 2 items. These ques-
behavior. Items were standardized, a summed tions addressed respondent’s self-reports of
score was calculated, and higher scores represent- spanking, slapping, and yelling behavior. The cor-
ed higher levels of harsh discipline (true for the relation between the two items was .31 (p , .01).
three age-based subsamples). Cronbach’s alpha Parent-adolescent conflict was measured using 10
was .89. Involvement was measured using three items that addressed how frequently the parent
items. These questions addressed the amount of and teen argued about various issues such as
time the respondent spent playing with or reading friends, the teen’s behavior, school, chores, and
to the child. Items were standardized, a summed getting along with family members. A mean score
score was calculated, and higher scores represent- was created, and high scores indicated more fre-
ed greater involvement with the child (true for the quent conflict. Cronbach’s alpha was .73. Involve-
three subsamples). Cronbach’s alpha was .64. A ment was measured using the same six items as
factor analysis that included the three harsh dis- those used in the families with a child aged 5
82 Journal of Marriage and Family

TABLE 1. INTERCORRELATIONS AND DESCRIPTIVE STATISTICS

Variables Parent Gender Child Gender Child Age Ethnic Background Poverty Status

Parent gender 07 01 203 211**

Child gender 202/00 06 02 03
Child age 00/204 202/202 201 04
Ethnic background 204/03 02/202 202/201 19***
Poverty status 201/03 07*/206 06/02 15***/35***
Child maladjustment 04/03 207*/01 08*/203 210***/202 01/09**
Marital conflict 201/202 208**/03 204/210** 204/201 04/203
Harsh discipline 12***/07* 212***/202 225***/214*** 202/201 04/07*
Parental involvement 29***/31*** 206/04 213***/224*** 212***/203 210**/206
Parental presence 204/02 08**/10** 244***/247*** 09**/22*** 05/08*
Parent-Adolescent Conflict NA/05 NA/204 NA/205 NA/212*** NA/205
M (SD) Child 2–4 .54 (.50) .48 (.50) 2.91 (.83) .17 (.37) .07 (.26)
M (SD) Child 5–11 .51 (.50) .50 (.50) 7.63 (1.94) .22 (.42) .07 (.26)
M (SD) Teen 12–18 .51 (.50) .48 (.50) 14.89 (1.83) .17 (.37) .07 (.25)
Note: Decimals are omitted from correlations. Correlations for families with child aged 2–4 are in upper triangle. Cor-
relations for families with child aged 5–11 are first estimates in the lower triangle. Correlations for families with a teen
aged 12–18 are second estimates in lower triangle. Parent gender is coded 0 5 father, 1 5 mother. Child gender is coded
0 5 son, 1 5 daughter. Parent ethnic background is coded 0 5 European American, 1 5 other. Poverty status is coded
0 5 not in poverty, 1 5 in poverty.
*p , .05. **p , .01. ***p , .001.

through 11. Cronbach’s alpha was .76. Parental sponse categories were: not true, sometimes true,
presence was measured using four items that ad- and often true (true for all three subsamples).
dressed how often the child was left alone at var- These items were taken from the Child Behavior
ious times during the day and night (same items Checklist, one of the most widely used and well-
as used in families with a child aged 5 through validated measures of child problem behaviors in
11, but one item was dropped because of low in- the field (Achenbach, 1991). A mean score was
ternal consistency). Cronbach’s alpha was .87. A calculated for each child, and a higher score rep-
factor analysis that included the 2 harsh discipline resented greater maladjustment (true for all three
items, the 10 parent-adolescent conflict items, the subsamples). Cronbach’s alpha was .53.
6 involvement items, and the 4 parental presence Global maladjustment was measured using
items indicated four factors, providing evidence of eight items in families with a target child aged 5
discriminant validity. The correlations among par- through 11. The questions addressed the parent’s
enting measures ranged from nonsignificant (par- perception of the child’s tendencies to lose his or
ent-adolescent conflict and parental presence—r her temper, be unhappy, be mean to others, try
5 .01, p 5 .81) to moderate in strength (parent new things (reverse), be cheerful and happy (re-
adolescent conflict and harsh discipline—r 5 .31, verse), be compliant (reverse), be sociable (re-
p , .001). Construct validity was evident for mea- verse), and be responsible (reverse). Cronbach’s
sures of harsh discipline, parental presence, and alpha was .60.
parent-adolescent conflict because of significant Global maladjustment was measured using 10
correlations with parental depressive affect, .15, items in families with a target child aged 12 to
.09, .27, respectively. 18. The questions addressed the parent’s percep-
tion of the child’s tendencies to lose his or her
Children’s maladjustment. Global maladjustment temper, be unhappy, be mean to others, be fearful
was measured using nine items in families with a or anxious, try new things (reverse), be cheerful
target child aged 2 through 4. The questions ad- and happy (reverse), be compliant (reverse), be
dressed the parent’s perception of the child’s ten- sociable (reverse), keep busy (reverse), and be re-
dencies to lose his or her temper, be fearful and sponsible (reverse). Cronbach’s alpha was .64.
anxious, be fussy or irritable, be mean to others,
try new things (reverse), keep busy (reverse), be Analytic Procedures
cheerful and happy (reverse), be compliant (re- There was little missing data (i.e., less than 5%
verse), and be sociable (reverse). The three re- for most items). Missing data were imputed using
Marital Conflict 83

TABLE 1. EXTENDED

Parent-
Parental Parental Adolescent
Child Maladjustment Marital Conflict Harsh Discipline Involvement Presence Conflict

204 01 11** 33*** NA NA

208* 01 212*** 05 NA NA
01 02 217*** 216*** NA NA
206 212** 212** 222*** NA NA
06 03 10** 218*** NA NA
21*** 29*** 216*** NA NA
21***/32*** 24*** 221*** NA NA
23***/28*** 14***/35*** 02 NA NA
214***/217*** 208**/202 09**/11*** NA NA
204/06 01/01 14***/06 09**/13***
NA/56*** NA/45*** NA/31*** NA/206 NA/01
1.50 (.24) .00 (.58) .00 (.91) .00 (.76) NA NA
1.42 (.27) .00 (.60) .00 (.94) .00 (.63) .00 (.74) NA
1.41 (.55) 2.01 (.55) 2.05 (.80) 2.05 (.67) 2.06 (.81) 1.63 (.53)

the expectation maximization (EM) method in with a child aged 5 through 11, the measures of
SPSS. This is a full information method of im- parental involvement and presence were included
puting missing values by iterating through the ex- in Block 1 as control variables, along with the
isting data and fitting the best values. This method background variables. Block 2 included the mar-
of handling missing values produces less bias in ital conflict variable.
the results than deleting cases or using the sample Within each group of families, additional re-
mean for imputation (Acock, 1997). Some of the gression analyses were conducted to ‘‘contextu-
parent involvement items had more missing data alize’’ the basic test of the spillover hypothesis.
in the sample of families with a target child aged This was done to see if the model fit equally well
2 through 4, and so, in addition to the imputation for (a) mothers’ and fathers’ reports, (b) focal sons
of values using EM, a mechanism variable mark- and daughters, (c) parents with European Ameri-
ing cases with missing values was included in the can ethnic backgrounds and parents with other
regression analysis for these families. The inclu- ethnic backgrounds, and (d) families living in
sion of the mechanism variable helps remove bias
poverty and families not living in poverty. Inter-
that might be present in the estimation of param-
action terms were created between each contex-
eters (Acock).
tual variable and each indicator of marital conflict
The spillover model was examined using hi-
erarchical regression. Two regression models were and parenting. These interaction terms were en-
fitted in each subsample of families. The first tered into the various regression equations as the
model provided estimates for the unmediated and last block or step.
mediated effects of marital conflict and for the
association between ineffective parenting and
children’s maladjustment. Block 1 included five RESULTS
variables: parent (respondent) gender, child gen-
der, race, family poverty status, and the missing
data mechanism variable (only in the analyses of The descriptive data and zero-order correlations
families with preschoolers). Block 2 included the are in Table 1. As hypothesized, marital conflict
marital conflict variable. Block 3 included the par- was associated positively with child and adoles-
enting variables. The second model provided es- cent maladjustment, positively with parents’ use
timates for the association between marital con- of harsh discipline, and inversely with parental in-
flict and parenting (the first leg of the spillover volvement in families with a target child younger
pathway). Block 1 included the control and mech- than 12. Harsh discipline was associated positive-
anism variables, as well as the parenting measures ly and parental involvement negatively with child
not being used as the dependent variable. For ex- and adolescent maladjustment. Parent-adolescent
ample, when the association between marital con- conflict was associated positively with marital
flict and harsh discipline was examined in families conflict and adolescent maladjustment.
84 Journal of Marriage and Family

FIGURE 1. MARITAL CONFLICT, INEFFECTIVE PARENTING, AND PRESCHOOLER’S MALADJUSTMENT IN TOTAL, EUROPEAN
AMERICAN, AND NON-WHITE SAMPLES, RESPECTIVELY

Note: Number in parentheses represents marital conflict estimate before parenting variables are entered.

Families with Preschoolers (2–4-year-olds) on preschooler’s maladjustment. In families with

a parent with an ethnic background other than Eu-
Spillover model. The association between marital
ropean American, the direct effect of marital con-
conflict and preschooler’s maladjustment was ex-
flict was moderate and neither harsh discipline nor
amined first. As hypothesized, marital conflict was
parental involvement intervened (see the third set
associated with greater maladjustment (Beta 5
of estimates in Figure 1). This group of families
.21, p , .001), controlling for background factors
included primarily Hispanic American or African
(see the first set of estimates in parentheses in Fig-
American parents.
ure 1). Harsh discipline (i.e., spanking or slap-
Parent’s gender contextualized (i.e., moderat-
ping) and parental involvement (i.e., playing with
ed) the model by shaping the association between
and reading to) partially mediated the association
between marital conflict and children’s maladjust- harsh discipline and preschooler’s maladjustment.
ment. Although the association between marital The relationship was statistically significant for
conflict and children’s maladjustment dropped in both mothers (b 5 .30, p , .001) and fathers (b
half (from .21 to .11), it remained statistically sig- 5 .16, p , .001), but was stronger for mothers.
nificant. About 11% of the variance in children’s Family poverty status contextualized the model by
maladjustment was accounted for by marital con- shaping the association between marital conflict
flict and ineffective parenting (net of background and parental involvement. The relationship was
factors). The results for the relationships between statistically significant in both groups of families
the background factors and child maladjustment but was stronger for families living in poverty (b
can be obtained from the first author. 5 2.51, p , .01) than for families not living in
poverty (b 5 2.23, p , .001).
Contextualizing the model. Except for modera-
tional effects of ethnicity, most of the interactions Alternative models. The redundancy model did
with background factors were not significant (i.e., not fit well for families with preschoolers. Marital
14 of 20 interaction terms were not significant, conflict and ineffective parenting were correlated,
70%). The model fit equally well for sons and but only in the .20s (about 5% shared variance).
daughters. The model was contextualized by the Both accounted for unique variance in preschool-
parent’s ethnic background, parent’s gender, and er’s maladjustment and therefore were tapping dif-
family poverty status. Ethnic background had the ferent familial characteristics. The independent-
strongest effect. In families with a European additive model did not fit the data well given the
American parent, the direct effect of marital con- results of the moderating effects of ethnic back-
flict became nonsignificant when harsh discipline ground. In European American families, parenting
and parental involvement were considered (see the was most salient, whereas in families with other
second set of estimates in Figure 1). Parenting ethnic backgrounds, marital conflict was most sa-
completely mediated the effects of marital conflict lient. There was some support for the interactive
Marital Conflict 85

model. Harsh discipline moderated the association ure 2). Harsh discipline and parental involvement
between marital conflict and preschooler’s mal- partially explained the association between marital
adjustment. Marital conflict was statistically sig- conflict and children’s maladjustment. Parental
nificant under most levels of harsh discipline but presence was not significant. Although marital
was slightly stronger when harsh discipline was conflict also was associated with children’s mal-
lower. Thus, physical punishment of the child adjustment through ineffective parenting, the di-
slightly buffered the deleterious effects of marital rect effect remained relatively unchanged
conflict, even though marital conflict and physical (dropped slightly from .20 to .16). About 12% of
punishment were positively correlated. Regardless the variance in children’s maladjustment was ac-
of the direction of effects, the moderating effects counted for by marital conflict and ineffective par-
were small. enting (net of background factors).
Thus, the spillover model fit these data the
best, but only in families with a European Amer- Contextualizing the model. The great majority of
ican parent. In this group of families, the associ- interaction terms with background factors were
ation between marital conflict and preschooler’s not significant (26 of 28 interaction terms were
maladjustment was mediated completely by inef- not significant, 93%). The model fit equally well
fective parenting (i.e., greater use of harsh disci- for mothers’ and fathers’ reports and for families
pline as measured by spanking or slapping and with sons and daughters. Two of the interaction
lower levels of parental involvement as measured terms were statistically significant—one between
by playing with or reading to). family poverty status and parental involvement in
predicting child maladjustment (the second leg of
Families With Children in Middle Childhood the pathway) and one between race and parental
(5–11-year-olds) involvement in predicting children’s maladjust-
ment (the second leg of the pathway). The asso-
Spillover model. The association between marital ciation between parental involvement and child
conflict and middle childhood children’s malad- maladjustment was statistically significant only
justment was examined next. Before including the for families not living in poverty (not poor b 5
parenting measures, marital conflict was associ- 2.19, p , .001; poor b 5 2.05, p . .05). The
ated with greater maladjustment (b 5 .20, p , association between parental involvement and
.001), controlling for background factors (see Fig- children’s maladjustment was statistically signifi-

FIGURE 2. MARITAL CONFLICT, INEFFECTIVE PARENTING, AND MALADJUSTMENT IN MIDDLE CHILDHOOD CHILDREN

Note: Number in parentheses represents marital conflict estimate before parenting variables are entered.
86 Journal of Marriage and Family

FIGURE 3. MARITAL CONFLICT, INEFFECTIVE PARENTING, AND MALADJUSTMENT IN ADOLESCENTS

Note: Number in parentheses represents marital conflict estimate before parenting variables are entered.

cant both for European American families and mi- when it was lower (b 5 .18, p , .001). Although
nority families (EA b 5 2.14, p , .001; non- this interaction provided some detail about the in-
White b 5 2.28, p , .001), but was stronger in terconnections between marital conflict and par-
non-White families. The pathway from marital enting, the effects were not strong. Marital conflict
conflict to harsh discipline to children’s malad- was associated with children’s maladjustment at
justment was similar for mothers and fathers, sons both levels of parental involvement, and neither
and daughters, families with varying ethnic back- harsh discipline nor parental presence interacted
grounds, and families living in and out of poverty. with marital conflict.
Thus, in families with a child in middle child-
Alternative models. The redundancy model did hood, there was evidence of (a) spillover effects,
not fit well for families with a target child in mid- in the form of indirect effects of ineffective par-
dle childhood. Marital conflict and ineffective par- enting; (b) independent-additive effects, in which
enting were correlated, but only in the .20s (about both marital conflict and ineffective parenting af-
5% shared variance). Each accounted for unique fected children’s maladjustment; and (c) interac-
variance in maladjustment and were therefore tap- tive effects, with the exacerbating effect of paren-
ping different familial characteristics. The inde- tal involvement being fairly weak.
pendent-additive model fit the data. Marital con-
flict and ineffective parenting each explained
Families With Adolescents (12–18-year-olds)
children’s maladjustment, and together they ex-
plained more variance in maladjustment than did Spillover model. The association between marital
either domain on its own. conflict and adolescent’s maladjustment was sig-
There also was some evidence of an interaction nificant (b 5 .32, p , .001; see Figure 3). Harsh
between marital conflict and ineffective parenting. discipline and parent-adolescent conflict fully me-
The association between marital conflict and chil- diated the association between marital conflict and
dren’s maladjustment was stronger when parental adolescent’s maladjustment. About 35% of the
involvement was higher (b 5 .24, p , .001) than variance in adolescent’s maladjustment was ac-
Marital Conflict 87

counted for by marital conflict and ineffective par- adolescent conflict to adolescent maladjustment
enting (net of background factors). was similar for mothers’ and fathers’ reports and
for poor and less poor families, but differed for
Contextualizing the model. Most of the interaction sons and daughters and for European American
terms with background factors were not signifi- and non–European American families. Adolescent
cant (28 of 36 terms, 78%). The direct path from gender moderated the association between marital
marital conflict to adolescent maladjustment was conflict and parent-adolescent conflict because it
similar for mothers and fathers, daughters and was stronger for daughters (b 5 .45, p , .001,
sons, families with parents of various ethnic back- controlling for other parenting variables) than for
grounds, and families living in and out of poverty. sons (b 5 .32, p , .001). This association also
The pathway from marital conflict to parent’s was stronger for European American families (b
harsh discipline to adolescent maladjustment dif- 5 .43, p , .001) than for minority families (b 5
fered by child gender, family poverty status, and .21, p , .001).
parent gender. The path differed for daughters and
sons because the association between marital con- Alternative models. The redundancy model did
flict and harsh discipline (the first leg and con- not fit well for families with adolescents. The cor-
trolling for the three other parenting variables) relations between marital conflict and ineffective
was stronger for sons (b 5 .33, p , .001) than parenting were low, and each accounted for
for daughters (b 5 .19, p , .001). The path dif- unique variance in adolescent maladjustment. The
fered for poor and less poor families because the independent-additive model fit the data. Parent-
association between marital conflict and harsh dis- adolescent conflict, harsh discipline, and involve-
cipline (the first leg) was statistically significant ment each were associated with adolescent mal-
in families not living in poverty (b 5 .30, p , adjustment and the measures of parenting
.001) but not for families living in poverty (b 5 explained an additional 25% of variance in mal-
.09, p . .05). The path differed for mothers and adjustment, above that explained by marital con-
fathers because the association between harsh dis- flict (i.e., 10%).
cipline and adolescent maladjustment (the second There also was some support for the interactive
leg) was significant for mothers (b 5 .23, p , model. The interactions between marital conflict
.001) but not for fathers (b 5 .02, p . .05). The and harsh discipline, marital conflict and parental
pathway was similar for families with different involvement, and marital conflict and parental
ethnic backgrounds. presence were not significant. The interaction
The pathway from marital conflict to parental between marital conflict and parent-adolescent
involvement to adolescent maladjustment differed conflict was significant, however. The association
by parent gender. The path from marital conflict between marital conflict and adolescent malad-
to involvement (the first leg) was statistically sig- justment was significant both when parent-adoles-
nificant for fathers (b 5 2.11, p , .001) but not cent conflict was above the mean (b 5 .21; p ,
for mothers (b 5 2.07, p . .05). Conversely, the .01) and lower (b 5 .11; p , .05) but was stron-
path from involvement to adolescent maladjust- ger when parent-adolescent conflict was higher.
ment (the second leg) was stronger for mothers (b In summary, the spillover and the independent-
5 2.25, p , .001) than for fathers (b 5 2.08, p additive models best fit the data on families with
, .05). The pathway was similar for daughters an adolescent. There was little support for the re-
and sons, for parents with different ethnic back- dundancy and interactive (particularly buffering)
grounds, and for families living in and out of pov- models. There was some support for the idea that
erty. particular aspects of ineffective parenting exacer-
The pathway from marital conflict to parental bate the association between marital conflict and
presence to adolescent maladjustment differed by adolescent maladjustment.
ethnic status. The path from marital conflict to
presence (the first leg) was statistically significant
DISCUSSION
for non-European American families (b 5 2.17,
p , .001) but not for European American families There are several useful findings in this study. The
(b 5 .02, p . .05). The pathway was similar for first is the confirmation that marital conflict is as-
daughters and sons, for families living in and out sociated with greater use of harsh discipline by
of poverty, and for mothers and fathers. parents, with reduced levels of parental involve-
The pathway from marital conflict to parent- ment, and with more frequent parent-adolescent
88 Journal of Marriage and Family

conflict. These, in turn, are associated with greater planations for these findings. First, because of the
levels of maladjustment in children and adoles- limited measurement of parenting in the NSFH,
cents. With only a few exceptions, the spillover we might not have included enough of the im-
from marital conflict to child maladjustment portant dimensions of parenting. Although the use
through ineffective parenting characterizes sons of harsh discipline and reduced parental involve-
and daughters, poor and less poor families, and ment are central components, we regret not having
from mothers’ and fathers’ perspectives. One of a measure of inconsistency. One of the hallmarks
the strengths of this research design is that we of emotional arousal that accompanies marital
were able to examine these spillover patterns conflict is that thinking and behavior become
across a wide age range of children, and the pat- more disorganized (Davies & Cummings, 1994).
tern fit across ages and developmental stages. This This disorganization makes it more difficult for a
pattern described families across the three age parent to respond consistently to a child’s misbe-
groups with the exception of non–European havior. Thus, a measure of parental inconsistency
American families with a target child aged 2 would have strengthened this study and might
through 4. Thus, one of the deleterious aspects of have resulted in a further reduction of the asso-
marital conflict, specifically disagreement and ver- ciation between marital conflict and children’s
bal aggression, is that parents also tend to spank, maladjustment. Given the strength of the associ-
slap, or yell at their children, as well as argue ation between marital conflict and parent-adoles-
more frequently with their child. This type of cent conflict in families with a target teen, it also
physical and verbal hostile behavior by parents, is unfortunate there wasn’t a measure of parent-
although clearly not abusive in most families, cov- child conflict in families with a target child youn-
aries with children’s and adolescents’ maladjust- ger than 12. A second reason for the existence of
ment. direct effects might be that marital conflict, in and
Another deleterious aspect of marital conflict of itself, might be associated with children’s ad-
is that it is associated with lower levels of parental justment difficulties. This might be through the
involvement. Parents are less likely to praise their direct modeling of aggressive ways of handling
children, read to them, play with them, and spend relational problems, through emotional and phys-
time with them engaging in relational and social iological arousal, and through increasing emotion-
activities. This reduced involvement also covaries al insecurity (Davies & Cummings, 1998).
with children’s and adolescents’ maladjustment. The existence of direct effects, unmediated by
The picture that emerges from these findings is parenting, was present in non–European American
consistent with an aversive chain of events de- families with a target child aged 2 through 4. So
scribed by Patterson (1982). As marital conflict little previous research has examined the patterns
escalates, parents might become increasingly ab- of marital conflict, parenting, and children’s mal-
sorbed by marital problems, less available to their adjustment through the lens of ethnicity that it is
children, more prone to forceful and punitive dis- difficult to interpret this finding. We conducted
ciplinary practices, and more likely to argue with follow-up analyses to determine whether this in-
their child. As a result of this withdrawal and co- teraction was influenced by differences in eco-
ercive, hostile behavior, youth might develop a nomic standing and it was not. Perhaps ethnic
pattern of behavior that is outwardly aggressive families are better at compartmentalizing their re-
and defiant. lational distress and not allowing the negative af-
A second important finding is that in families fect from their marriage influence their parenting.
with a target child aged 2 through 11, parenting This is an area in need of further research.
only partially explained the association between The spillover model was similar for mothers
marital conflict and children’s maladjustment. and fathers in families with a target child aged 2
Crockenberg and Covey (1991) believed that it through 11. However, the model differed for
was unclear whether marital conflict still would mothers and fathers with a target child in adoles-
be associated with children’s maladjustment when cence because the inverse association between
parenting effects were simultaneously considered. marital conflict and parental involvement was sta-
In this study, direct effects remained. Emery et al., tistically significant only for fathers. Previous lit-
(1992) contended that the effects of marital con- erature reviews have suggested that marital con-
flict cannot be reduced to parenting disruptions. flict might affect fathering more than mothering
The findings from this study support their conten- (Crockenberg & Covey, 1991; Krishnakumar &
tion. There seem to be at least two plausible ex- Buehler, 2000). The thinking behind this possible
Marital Conflict 89

gender difference is that mothering might be less ager are able to compartmentalize their marital
contextual than fathering. Recent literature rein- conflict.
forces this notion. In their narrative review of the The linkage through reduced parental involve-
fathering literature, Doherty, Kouneski, and Erick- ment also did not differ by poverty status except
son (1998) concluded that fathering is influenced for families with a target child aged 5 through 11.
by family and community factors to a greater de- In these families, the pathway through reduced pa-
gree than is mothering. These authors provided rental involvement existed only for families not
evidence that residential father involvement is living in poverty. Perhaps this was a function of
contingent on several external factors including number of hours employed, with families not liv-
mother’s attitudes toward, expectations of, and ing in poverty working more in the face of marital
support for the father. Thus, when marital tension conflict. Increased employment activity might oc-
is high, fathers might feel unsupported by their cur to avoid one’s spouse or to bolster economic
spouse and consequently disengage from active well-being in the case of a future separation. Fu-
involvement with their children. The restriction of ture longitudinal research needs to include em-
this finding to adolescents suggests that the gen- ployment activity in the model because it is an
eral pattern of reduced father involvement during alternative use of time, when considering spend-
adolescence (Pleck, 1997) might be triggered, in ing time with children.
part, by marital strain. A final point of discussion addresses the utility
The pathway from marital conflict to child of the spillover model as compared with other
maladjustment through harsh discipline existed for ways in which marital conflict, ineffective parent-
sons and daughters from ages 2 through 18. The ing, and children’s maladjustment can interrelate.
significant indirect effects through parental in- The results of this study suggest that the spillover
volvement also were similar for sons and daugh- model is a reasonable way to model interconnec-
ters in families with a child aged 2 through 11. tions. For the most part, marital conflict is asso-
The path through parent-adolescent conflict was ciated with higher levels of ineffective parenting,
stronger for daughters than sons, however. At a and this pattern is robust when various back-
time when identity issues and peer relationships ground characteristics are taken into consider-
become more salient in the child’s life, marital ation.
conflict placed daughters in greater jeopardy. This Because ineffective parenting does not fully
is consistent with Gilligan’s (1982) theoretical account for the association between marital con-
suggestions that daughters might be more vulner- flict and children’s maladjustment, the indepen-
able to family stressors (i.e., marital and parent- dent-additive model also represented the data fair-
adolescent conflict) because they are socialized to ly well. Both marital conflict and ineffective
caretake and safeguard family relationships. parenting were associated with children’s adjust-
In addition to the differences already noted be- ment difficulties, and each contributed uniquely to
tween European American and non–European the explanation of these difficulties. Neither the
American families with a target child aged 2 redundant model nor the buffering model fit the
through 4, the leg between marital conflict and data well.
parental presence was significant in non–Europe- Although this study of the interconnections
an American families with an adolescent but not among marital conflict, parenting, and children’s
in European American families. Marital conflict maladjustment in two-parent families overcame
was associated with reduced parental presence. In many of the methodological limitations of previ-
addition to reduced time spent together, marital ous research, it, too, had limitations. The internal
conflict is a risk factor for minority youth because consistency reliabilities of child maladjustment
it is associated with parents not being at home. were low. Although these measures of maladjust-
The pathways from marital conflict to chil- ment have been used in previous research using
dren’s maladjustment through harsh discipline the NSFH and have demonstrated evidence of
were similar for families living in and out of pov- construct validity (Acock & Demo, 1994), they
erty except in families with a target teen. In these contain random error that might serve to depress
families, the link between marital conflict and the covariances. Low reliability on the dependent
harsh discipline only existed in families with in- measure does not create instability in the predictor
comes above the poverty line. If this finding rep- estimations (as low reliability on the independent
licates in other studies, additional research is variables can) but reduces the amount of variance
needed to explain how poor families with a teen- explained (Maruyama, 1998). Thus, the variance
90 Journal of Marriage and Family

explained in children’s (13% and 12%) and ado- REFERENCES

lescents’ (36%) maladjustment is underestimated. Achenbach, T. M. (1991). Manual for the child behavior
The cross-sectional design also presents limi- checklist/4–18 and 1991 profile. Burlington: Univer-
tations in making causal inferences. The signifi- sity of Vermont, Department of Psychiatry.
cant patterns among correlations can be examined, Acock, A. (1997). Working with missing data. Family
but the time ordering among variables is not clear. Science Review, 10, 76–102.
Acock, A. C., & Demo, D. H. (1999). Dimensions of
Thus, in terms of the model tested in the present family conflict and their influence on child and ado-
study, we cannot be certain that marital conflict lescent adjustment. Sociological Inquiry, 69, 641–
preceded the ineffective parenting nor that the in- 658.
effective parenting preceded the children’s diffi- Acock, A. C., & Demo, D. H. (1994). Family diversity
culties. Although longitudinal evidence is scant, and well-being. Thousand Oaks, CA: Sage.
Almeida, D. M., Wethington, E., & Chandler, A. L.
studies that have examined the causal ordering be- (1999). Daily transmission of tensions between mar-
tween marital conflict and parenting, as well as ital dyads and parent-child dyads. Journal of Mar-
marital conflict and child maladjustment, have re- riage and the Family, 61, 49–61.
ported that marital conflict seems to proceed each Anderson, E. R., Lindner, M. S., & Bennion, L. D.
(1992). The effect of family relationships on adoles-
(Acock & Demo, 1999; Fincham, Grych, & Os- cent development during family reorganization. In E.
borne, 1994; Floyd, Gilliom, & Costigan, 1998). M. Hetherington & W. G. Clingempeel (Eds.), Cop-
Again not tested thoroughly, there is at least some ing with marital transitions: A family systems per-
evidence that child maladjustment proceeds inef- spective (pp. 178–199). Chicago: University of Chi-
fective parenting in some families (particularly for cago Press.
Belsky, J. (1984). The determinants of parenting: A pro-
sons and externalizing problems; Lytton, 1990; cess model. Child Development, 55, 83–96.
Sampson & Laub, 1994; Simons et al., 1994; Brody, G. H., Stoneman, Z., & Flor, D. (1995). Linking
Zahn-Waxler, 1996). Clearly, three-wave longitu- family processes and academic competence among
dinal research is needed to ensure that the vari- rural African American youths. Journal of Marriage
and the Family, 57, 567–579.
ables are time ordered in such a way that the caus- Buehler, C., Anthony, C., Krishnakumar, A., Stone, G.,
al ordering of the spillover effects can be Gerard, J., & Pemberton, S. (1997). Interparental con-
examined more carefully. flict and youth problem behaviors: A meta-analysis.
This study makes an important contribution to Journal of Child and Family Studies, 6, 233–247.
the literature on marital conflict and youth mal- Cowan, C. P., Cowan, P. A., Heming, G., & Miller, N.
B. (1991). Becoming a family: Marriage, parenting,
adjustment by delineating ways in which parent- and child development. In P. A. Cowan & M. Heth-
ing is susceptible to the influence of negative mar- erington (Eds.), Family Transitions (pp. 79–109).
ital interaction patterns. Although the marital Hillsdale, NJ: Erlbaum.
relationship has long been recognized theoretical- Crockenberg, S. B., & Covey, S. L. (1991). Marital con-
flict and externalizing behavior in children. In D. Cic-
ly as a primary determinant of parenting quality chetti & S. Toth (Eds.), Rochester Symposium on De-
and a principal support system to parents (Belsky, velopmental Psychopathology: Vol. 3. Research and
1984), research has been limited by insufficient clinical contributions to a theory of developmental
tests of the spillover hypothesis. Findings from psychopathology (pp. 235–360). Rochester, NY: Uni-
this study can be used to inform clinicians and versity of Rochester Press.
Cummings, E. M., & Davies, P. (1994). Children and
family life educators about the specific ways in marital conflict: The impact of family dispute and res-
which hostile marital interactions can compromise olution. New York: Guilford.
the quality of parental caregiving and, ultimately, Davies, P. T., & Cummings, E. M. (1998). Exploring
children’s developmental growth. This informa- children’s emotional security as a mediator of the link
between marital relations and child adjustment. Child
tion has important implications for direct inter- Development, 69, 124–139.
vention and the planning of effective programs for Davies, P. T., & Cummings, E. M. (1994). Marital con-
populations at risk. Future research can supple- flict and child adjustment: An emotional security hy-
ment this information by examining the role of pothesis. Psychological Bulletin, 116, 387–411.
ethnicity in the relationships among marital con- Doherty, W. J., Kouneski, E. F., & Erickson, M. F.
(1998). Responsible fathering: An overview and con-
flict, parenting quality, and youth maladjustment ceptual framework. Journal of Marriage and the
and by examining these associations over time and Family, 60, 277–292.
across the life span. Emery, R. E., Fincham, F. D., & Cummings, E. M.
(1992). Parenting in context: Systemic thinking about
parental conflict and its influence on children. Journal
NOTE of Consulting and Clinical Psychology, 60, 909–912.
We thank John G. Orme for his methodological advice Emery, R. E., & Tuer, M. (1993). Parenting and the
regarding the statistical effects of low reliability. marital relationship. In T. Luster & L. Okagaki (Eds.),
Marital Conflict 91

Parenting: An ecological perspective. (pp. 121–148). izing behaviors. Developmental Psychology, 29, 940–
Hillsdale, NJ: Erlbaum. 950.
Engfer, A. (1988). The interrelatedness of marriage and Kim, J. E., Hetherington, E. M., & Reiss, D. (1999).
the mother-child relationship. In R. A. Hinde & J. Associations among family relationships, antisocial
Stevenson-Hinde (Eds.), Relationships within fami- peers, and adolescents’ externalizing behaviors: Gen-
lies: Mutual influences (pp. 104–118). Oxford, U.K.: der and family type differences. Child Development,
Clarendon. 70, 1209–1230.
Erel, O., & Burman, B. (1995). Interrelatedness of mar- Krishnakumar, A., & Buehler, C. (2000). Interparental
ital relations and parent-child relations: A meta-ana- conflict and parenting behaviors: A meta-analytic re-
lytic review. Psychological Bulletin, 118, 106–132. view. Family Relations, 49, 25–44.
Fauber, R., Forehand, R., Thomas, A., & Wierson, M. Linver, M. R., & Silverberg, S. B. (1995). Parenting as
(1990). A mediational model of the impact of marital a multidimensional construct: Differential prediction
conflict on adolescent adjustment in intact and di- of adolescents’ sense of self and engagement in prob-
vorced families: The role of disrupted parenting. lem behavior. International Journal of Adolescent
Child Development, 61, 1112–1123. Medicine and Health, 8, 29–40.
Fauber, R. L., & Long, N. (1991). Children in context: Linver, M. R., & Silverberg, S. B. (1997). Maternal pre-
The role of the family in child psychotherapy. Jour- dictors of early adolescent achievement-related out-
nal of Consulting and Clinical Psychology, 59, 813– comes: Adolescent gender as a moderator. Journal of
820. Early Adolescence, 17, 294–318.
Fincham, F. D. (1994). Understanding the association Lytton, H. (1990). Child and parent effects in boys’ con-
between marital conflict and child adjustment: Over- duct disorder: A reinterpretation. Developmental Psy-
view. Journal of Family Psychology, 2, 123–127. chology, 26, 683–697.
Fincham, F. D., Grych, J. H., & Osborne, L. N. (1994). Lytton, H., & Romney, D. (1991). Parents’ differential
Does marital conflict cause child maladjustment? Di- socialization of boys and girls: A meta analysis. Psy-
rections and challenges for longitudinal research. chological Bulletin, 109, 267–296.
Journal of Family Psychology, 8, 128–140. Mann, B. J., & MacKenzie, E. P. (1996). Pathways
Floyd, F. J., Gilliom, L. A., & Costigan, C. L. (1998). among marital functioning, parental behaviors, and
Marriage and the parenting alliance: Longitudinal child behavior problems in school-age boys. Journal
prediction of change in parenting perceptions and be- of Clinical Child Psychology, 25, 183–191.
haviors. Child Development, 69, 1461–1479. Maruyama, G. M. (1998). Basics of structural equation
Gilligan, C. (1982). In a different voice: Psychological modeling. Thousand Oaks, CA: Sage.
theory and women’s development. Cambridge, MA: Miller, N. B., Cowan, P. A., Cowan, C. P., Hetherington,
Harvard University Press. E. M., & Clingempeel, W. G. (1993). Externalizing
Gonzales, N. A., Pitts, S. C., Hill, N. E., & Roosa, M. in preschoolers and early adolescents: A cross-study
W. (2000). A mediational model of the impact of in- replication of a family model. Developmental Psy-
terparental conflict on child adjustment in a multieth- chology, 29, 3–18.
nic, low-income sample. Journal of Family Psychol- O’Brien, M., Margolin, G., & John, R. S. (1995). Re-
ogy, 14, 365–379. lation among marital conflict, child coping, and child
Gottman, J. M., & Katz, L. F. (1989). Effects of marital adjustment. Journal of Clinical Child Psychology, 24,
discord on young children’s peer interaction and 346–361.
health. Developmental Psychology, 25, 373–381. Osborne, L. N., & Fincham, F. D. (1996). Marital con-
flict, parent-child relationships, and child adjustment:
Grych, J. H., & Fincham, F. D. (1990). Marital conflict
Does gender matter? Merrill-Palmer Quarterly, 42,
and children’s adjustment: A cognitive-contextual
48–75.
framework. Psychological Bulletin, 108, 267–290. Patterson, G. R. (1982). Coercive family process. Eu-
Grych, J. H., Seid, M., & Fincham, F. D. (1992). As- gene, OR: Castalia.
sessing marital conflict from the child’s perspective: Peterson, G. W., Bodman, D. A., Bush, K., & Madden-
The children’s perception of interparental conflict Derdich, D. (2000). Gender and parent-child relation-
scale. Child Development, 63, 558–572. ship. In D. H. Demo, K. R. Allen, & M. A. Fine
Harold, G. T., & Conger, R. D. (1997). Marital conflict (Eds.), Handbook of family diversity (pp. 82–104).
and adolescent distress: The role of adolescent aware- New York: Oxford University Press.
ness. Child Development, 68, 333–350. Peterson, J. L., & Zill, N. (1986). Marital disruption,
Harold, G. T., Fincham, F. D., Osborne, L. N., & Conger, parent-child relationships, and behavior problems in
R. D. (1997). Mom and dad are at it again: Adoles- children. Journal of Marriage and the Family, 48,
cent perceptions of marital conflict and adolescent 295–307.
distress. Developmental Psychology, 33, 333–350. Pleck, J. H. (1997). Paternal involvement: Levels,
Harrist, A. W., & Ainslie, R. C. (1998). Marital discord sources, and consequences. In M. E. Lamb (Ed.), The
and child behavior problems: Parent-child relation- role of the father in child development (pp. 66–103).
ship quality and child interpersonal awareness as me- New York: Wiley.
diators. Journal of Family Issues, 19, 140–163. Reid, W. J., & Crisafulli, A. (1990). Marital discord and
Jenkins, J. M., Smith, M. A., & Graham, P. J. (1989). child behavior problems: A meta-analysis. Journal of
Coping with parental quarrels. Journal of American Abnormal Child Psychology, 18, 105–117.
Academy of Child and Adolescent Psychiatry, 28, Repetti, R. L. (1987). Links between work and family
182–189. role. In S. Oskamp (Ed.), Family processes and prob-
Katz, L. F., & Gottman, J. M. (1993). Patterns of marital lems: Social psychological aspects (pp. 98–127).
conflict predict children’s internalizing and external- Newbury Park, CA: Sage.
92 Journal of Marriage and Family

Rothbaum, F., & Weisz, J. R. (1994). Parental caregiv- Stone, G., Buehler, C., & Barber, B. K. (2002). Inter-
ing and child externalizing behavior in nonclinical parental conflict, parental psychological control, and
samples: A meta-analysis. Psychological Bulletin, youth problem behaviors. In B. K. Barber (Ed.), In-
116, 55–74. trusive parenting: How psychological control affects
Sampson, R. J., & Laub, J. H. (1994). Urban poverty children and adolescents (pp. 53–95). Washington,
and the family context of delinquency: A new look DC: American Psyhological Association.
at structure and process in a classic study. Child De- Sweet, J., Bumpass, L., & Call, V. (1992). National Sur-
velopment, 65, 523–540. vey of Families and Households 1988. Madison, WI:
Sanders, M. R., Nicholson, J. M., & Floyd, F. J. (1997). Center for Demography an Ecology, University of
Couples’ relationships and children. In W. K. Halford Wisconsin.
& H. J. Markman (Eds.), Clinical handbook of mar- Vandewater, E. A., & Lansford, J. E. (1998). Influence
riage and couples interventions (pp. 225–253). New
of family structure and parental conflict on children’s
York: Wiley.
Simons, R. L., and associates. (1996). Understanding well-being. Family Relations, 47, 323–330.
differences between divorced and intact families. Wilson, B. J., & Gottman, J. M. (1995). Marital inter-
Thousand Oaks, CA: Sage. action and parenting. In M. H. Bornstein (Ed.), Hand-
Simons, R. L., Whitbeck, L. B., Beaman, J., & Conger, book of Parenting: Volume 4, Applied and practical
R. D. (1994). The impact of mothers’ parenting, in- parenting (pp. 33–55). Mahwah, NJ: Erlbaum.
volvement by nonresidential fathers, and parental Zahn-Waxler, C. (1996). Environment, biology, and cul-
conflict on adjustment of adolescent children. Journal ture: Implications for adolescent development. De-
of Marriage and the Family, 56, 356–374. velopmental Psychology, 32, 571–573.