ArticlePDF Available

Assessing the Links Between Interparental Conflict and Child Adjustment: The Conflicts and Problem-Solving Scales

Authors:

Abstract and Figures

Two hundred seventy-three couples completed the Conflicts and Problem-Solving Scales (CPS), which assess dimensions of interparental conflict that affect children, including frequency, severity, resolution, and efficacy. Factor analyses revealed reliable scales for conflict strategies involving cooperation, avoidance, stalemate, physical aggression, verbal aggression, and child involvement. The CPS correlated with other measures of marital conflict, and gender differences in partners' conflict strategies were found. One hundred sixteen of the wives and 79 of the husbands participated in a larger study involving their school-age child. Parents' and children's descriptions of interparental conflict agreed with one another. There were gender differences in the ways in which the CPS scales were related to parents' and children's reports of child symptoms. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Content may be subject to copyright.
Journal of Family Psychology
1996,
Vol. 10, No. 4, 454-473Copyright 1996 by the American Psychological Association, Inc.
0893-320(V96/$3.00
Assessing the Links Between Interparental
Conflict and Child Adjustment:
The Conflicts and Problem-Solving Scales
Patricia K. Kerig
Simon Fraser University
Two hundred seventy-three couples completed the Conflicts and Problem-Solving
Scales (CPS), which assess dimensions of interparental conflict that affect children,
including frequency, severity, resolution, and efficacy. Factor analyses revealed reli-
able scales for conflict strategies involving cooperation, avoidance, stalemate, physical
aggression, verbal aggression, and child involvement. The CPS correlated with other
measures of marital conflict, and gender differences in partners' conflict strategies
were found. One hundred sixteen of the wives and 79 of the husbands participated in
a larger study involving their school-age child. Parents' and children's descriptions of
interparental conflict agreed with one another. There were gender differences in the
ways in which the CPS scales were related to parents' and children's reports of child
symptoms.
Exposure to interparental conflict is a preva-
lent source of stress for children. Conflict
within a marriage has been shown to have neg-
ative effects on parenting (Belsky, Rovine, &
Fish, 1989; C. P. Cowan, Cowan, Heming, &
Miller,
1991;
Cox, Owen, Lewis, & Henderson,
1989;
Kerig, Cowan, & Cowan, 1993; Pratt,
Kerig, Cowan, & Cowan, 1992) and to be pre-
dictive of problems in child adjustment (Amato
& Keith, 1991; Crockenberg & Covey, 1991;
Jouriles, Murphy, & O'Leary, 1989; Jouriles,
This research was funded by Grant 410-94-1547
from the Social Science and Humanities Research
Council of Canada. Portions of these data were pre-
sented at the biennial meetings of the Society for
Research in Child Development, March 1995, Indi-
anapolis, Indiana.
I thank Corina Brown for her help with all phases
of the research, the families who participated, and the
following students who assisted in data collection,
entry, and coding: Surinder Antal, Mandeep Arneja,
Karen Bentley, Ken Brown, Sally Caron, Sandra
Chopping, Wendy Chua, Linda Duarte, Cheyene
Dyer, Janice Ebenstiner, Rachel French, Suzanne
Hutchinson, Mary Ann Jean, Greg Klassen, Lin Lim,
Riyaz Motan, and Renee Patenaude.
Correspondence concerning this article should be
addressed to Patricia K. Kerig, Department of Psy-
chology, Simon Fraser University, Burnaby, British
Columbia, Canada V5A 1S6. Electronic mail may be
sent via Internet to pkerig@arts.sfu.ca.
Pfiffner, & O'Leary, 1988; Katz & Gottman,
1993;
Porter & O'Leary, 1980). However, re-
searchers are increasingly turning their attention
to studying the mechanisms by which marital
conflict affects child development. Marriages
may not differ so much in whether or not con-
flicts occur but, rather, on such dimensions as
their severity, frequency, and resolution (Grych
& Fincham, 1990). Although severe couple
conflicts may place families at risk for other
major stressors, such as family violence or mar-
ital dissolution, conflict can also be viewed as a
normative part of family life (Emery, 1992;
Fitzpatrick, 1988; Katz, Kramer, & Gottman,
1992).
Therefore, it is important to understand
how even well-functioning families negotiate
their differences.
Understanding the effects of interparental
conflict on children has been limited by the
available measures of the marital relationship,
which tend to focus on unidimensional ratings
of marital quality or frequency of conflict. The
lack of specificity in the commonly used mea-
sures of marital relationships has been recently
criticized by a number of researchers (Chris-
tensen & Heavey, 1990; Grych, Seid, & Fin-
cham, 1992; Jenkins & Smith, 1991; Jouriles,
Murphy, et al., 1991). Although advances have
been made in discriminating patterns of couple
conflict predictive of marital adjustment (Chris-
454
INTERPARENTAL CONFLICT AND CHILD ADJUSTMENT455
tensen & Heavey, 1990; Noller & White, 1990),
the need remains for more attention to factors
that might moderate the influence of marital
distress on parenting and child development. An
important contribution to this research endeavor
was Grych et al.'s (1992) development of the
Children's Perceptions of Interparental Conflict
Scale, which assesses a number of qualities of
interparental disagreements from the child's
perspective. However, this measure has not yet
been complemented by an adult self-report mea-
sure with the same degree of specificity regard-
ing the aspects of marital conflict that affect
parenting. Parents and children may have dif-
ferent, but equally valid, perceptions of
the
fam-
ily system of relationships (Kerig, 1995b), and
it is therefore useful to gain data from multiple
perspectives in the family.
The purpose of the present study was to fill
this gap in the existing methodology by devel-
oping a new measure of couple conflict strate-
gies.
The Conflicts and Problem-Solving Scales
(CPS) were developed to assess the major di-
mensions of interparental conflict identified in
the literature: their Severity and frequency, the
content of disagreements, whether or not con-
flicts are satisfactorily resolved, the perceived
effectiveness of partners' problem-solving abil-
ities,
and the various conflict strategies partners
use in their attempts to resolve conflicts.
Dimensions of Interparental Conflict
Severity and Frequency
The more that interparental conflicts are dis-
ruptive of family functioning, the more likely it
is that children will perceive them to be distress-
ing (Grych & Fincham, 1990), and the more
likely it is that parents' quarrels will interfere
with children's ability to rely on their family as
a safe emotional environment in which their
needs can be met (Davies & Cummings, 1994).
Consistent with this, empirical evidence con-
firms that children who perceive their parents'
fights as frequent and intense also are higher in
symptoms and behavioral problems (Grych &
Fincham, 1993).
Content of Disagreements
The measures of marital quality and conflict
most frequently used in research on marriage
and parenting include the Marital Agendas Pro-
tocol (Notarius & Vanzetti, 1984), the Marital
Adjustment Test (Locke & Wallace, 1959), the
Dyadic Adjustment Scale (DAS; Spanier,
1976),
and PREPARE (Flowers & Olson,
1986).
Each measure presents a list of common
issues about which partners are asked to rate
their frequency of disagreement. Although their
structure is similar, the content of these mea-
sures is widely divergent. Among the 21 topics
listed by these measures altogether, only 6 ap-
pear on three of the four questionnaires. The
lack of correspondence among the topics cov-
ered by these measures is problematic in that
couples may quarrel with different levels of
frequency and urgency about some issues than
others.
In addition, one centrally important area of
potential disagreement between parents con-
cerns content related to children and child-
rearing issues. It has been posited that quarrels
between parents that are directly related to chil-
dren, such as disagreements about child rearing,
will have a more negative impact on parenting
and child development (Dadds & Powell, 1991;
Forehand & McCombs, 1989; Grych & Fin-
cham, 1990; Jouriles, Murphy, et al., 1991).
However, child content is included in only two
of the measures of marital discord most widely
used in research. Therefore, there is a need for
a measure of marital conflict that is comprehen-
sive in its inquiry as to the kinds of issues over
which couples might quarrel and that, in partic-
ular, includes questions as to how frequently
children and child rearing are topics of disagree-
ment.
Conflict Resolution and Efficacy
Another aspect of interparental conflict that
affects children is the degree to which quarrels
are successfully resolved (E. M. Cummings,
Ballard, El-Sheikh, & Lake, 1991; Grych &
Fincham, 1990). Children perceive resolved an-
ger as a less negative event (E. M. Cummings,
Vogel, Cummings, & El-Sheikh, 1989), and
resolution can ameliorate the effects of observ-
ing interadult aggression (E. M. Cummings et
al.,
1991). The CPS provides for an assessment
of the perceived outcome of conflict resolution
strategies based on an adaptation of
Rands,
Lev-
inger, and Mellinger's (1981) measure. As de-
rived from the Marital Agendas Protocol (No-
456KERIG
tarius & Vanzetti, 1984), the CPS also allows
for an assessment of relational efficacy and the
effectiveness of problem-solving abilities (mea-
sured by the proportion of quarrels that partners
satisfactorily resolve).
Conflict Strategies
The measure of conflict strategies currently
used most widely by researchers in the area of
marriage and parenting, the Conflict Tactics
Scales (CTS; Straus, 1979), tends to be heavily
weighted on the violent and dysfunctional end
of the conflict spectrum, and there are only a
small number of items available to construct
additional subscales that would allow for finer
discriminations among nonaggressive strategies
(cf. Margolin, 1990). Although exposure to ver-
bally and physically abusive interparental con-
flicts is associated with distress and maladjust-
ment in children (Fantuzzo et al., 1991; Holden
& Ritchie, 1991; Jouriles et al., 1989), it has
also been proposed that children may benefit
from observing their parents engage in conflict
resolution involving mutually respectful prob-
lem solving (Davies & Cummings, 1994; Em-
ery, 1992; Grych & Fincham, 1993). Therefore,
a broader sampling of prosocial strategies is
needed to test the hypothesis that some forms of
nonaggressive marital conflict may be construc-
tive for couples and their children.
In addition, assessment of a wider range of
conflict strategies may provide an opportunity
to examine gender differences in the ways that
men and women respond to marital conflict,
differences that are becoming of increasing in-
terest in research on marriage and parenting
(P.
A. Cowan, Cowan, & Kerig,
1993;
Howes &
Markman, 1989; Jouriles & LeCompte, 1991;
Noller & Fitzpatrick, 1991; Parke & Tinsley,
1987).
Husbands and wives differ in their pre-
ferred ways of handling conflict, and these dif-
ferences may be greater in distressed marriages
(Markman, Silvern, Clements, & Kraft-Hanak,
1993).
Whereas women are more likely to en-
gage directly in conflict with their spouses, men
are described as withdrawing or avoiding con-
flict (Christensen & Heavey, 1990; Gottman &
Levenson, 1988). Furthermore, as Miller, Dana-
her, and Forbes (1986) have pointed out, most
researchers have defined conflict as involving
attempts to influence another person to accept
one's own point of view. An alternate perspec-
tive,
in which conflict is defined as a challenge
to the maintenance of interpersonal harmony,
calls for strategies for mitigating negative affect
in oneself and others. Conflict mitigation is a
strategy more frequently observed in girls dur-
ing peer conflicts (Miller et al., 1986) and in
mothers and daughters during family interac-
tions (Vuchinich, Emery, & Cassidy, 1988).
However, conflict mitigation strategies have not
been assessed by the majority of couple conflict
measures used to date.
Therefore, as a means of affording an oppor-
tunity to explore hypotheses regarding con-
structive versus aggressive conflicts, as well as
gender differences, a measure was constructed
to provide a more detailed typology of nonvio-
lent strategies for marital problem solving.
These strategies were derived from a review of
the literature on marital conflict and conflict
resolution, including Christensen and Heavey
(1990),
Gottman (1995), Noller and White
(1990),
and Vuchinich (1987). In particular,
Pruitt and Rubin's (1986) differentiation of
strategies of avoidance, yielding, contending,
compromise, and collaboration was used. Ac-
cording to Pruitt and Rubin (1986), avoidance
allows one to escape from conflict, whereas
yielding represents a nonassertive strategy de-
signed to preserve harmony at all costs. Con-
tending involves attempting to control another
person to get one's individual needs met,
whereas compromise involves meeting the
other person halfway. Such attempts are to be
distinguished from true collaboration, which in-
volves joint problem solving to find a solution
that takes into account both partners' needs.
However, the typologies just described fail to
include abusive or violent conflict strategies,
which have important consequences for the de-
velopment of children who are exposed to them
(E.
M. Cummings, Zahn-Waxier, & Radke-
Yarrow, 1981; Sternberg et al., 1993). There-
fore,
scales regarding verbal aggression and
physical aggression were included in the present
measure.
Child Involvement in Interparental
Conflict
Conflict between parents can be expressed in
a number of different ways within the family,
and an understanding of how children are af-
fected by interparental discord may be im-
INTERPARENTAL CONFLICT AND CHILD ADJUSTMENT457
proved by attending to the degree to which
children are involved in their parents' argu-
ments. Children have been found to show more
distress when they are directly exposed to in-
terparental quarreling (J. S. Cummings, Pelle-
grini, Notarius, & Cummings, 1989; Forehand
& McCombs, 1989; Grych & Fincham, 1993;
O'Brien, Margolin, John, & Krueger, 1991;
Porter & O'Leary, 1980) or when the content of
interparental disagreements concerns children
(Jouriles, Murphy, et al., 1991) and child-rear-
ing issues (Jenkins & Smith, 1991; Snyder,
1979).
However, these constructs do not en-
compass the entirety of ways that marital con-
flict may affect parent-child relationships.
Children may not only observe their parents
quarreling or hear their parents fighting about
them but may also become directly involved in
interparental conflicts, such as by attempting to
intervene (E. M. Cummings, Iannotti, & Zahn-
Waxler, 1985; O'Brien et al., 1991; Vuchinich
et al., 1988) or by becoming "triangulated"
within the marital relationship (Minuchin, 1974;
Westerman, 1987). Triangulation may take a
variety of forms (Engfer, 1988; Kerig, 1995b;
Minuchin, 1974). For example, a child may
develop a special relationship or coalition with
one parent, play the role of go-between in at-
tempting to mediate parental disputes, or be-
come a scapegoat for parents who are attempt-
ing to circumvent or displace their marital strife.
Children's experiences of loyalty conflicts or
being "caught in the middle" between parents
have been studied in the context of divorce
(Buchanan, Maccoby, & Dornbusch, 1991; Ca-
mara & Resnick, 1988; Johnston, Campbell, &
Mayes, 1985) as well as intact families (Chris-
tensen & Margolin, 1988; Gilbert, Christensen,
& Margolin, 1984; Jouriles, Bourg, & Farris,
1991;
Kerig, 1995b; Vuchinich et al., 1988).
Although Grych et al.'s (1992) measure of
the child's perspective on interparental conflict
includes a subscale of triangulation, there
is a need for a measure that assesses this phe-
nomenon from the parents' perspective. The
O'Leary-Porter Scale (OPS; Porter & O'Leary,
1980) has been widely used as a parent-report
measure of children's exposure to interparental
conflict. However, this scale does not reflect the
child's active involvement and intervention in
parental conflicts in a manner that parallels the
triangulation subscale in Grych et al.'s (1992)
child measure, nor does it capture the different
forms that triangulation might take.
The Present Study
In sum, the purpose of the present study was
to develop a measure of interparental conflict
that would be sensitive to the different dimen-
sions of conflict that might affect parenting and
child development. The factor structure, reli-
ability, and validity of this new measure were
assessed, as well as its utility in predicting mar-
ital quality and satisfaction with conflict reso-
lution. A subset of couples with school-age
children was included to assess the correspon-
dence between parents' and children's percep-
tions of interparental quarreling and to assess
the relationship between the CPS and children's
adjustment.
Method
Participants
Participants in this research were 273 couples re-
cruited through letters sent to day-care centers, pre-
schools, and couple relationship and parenting class-
es;
fliers posted at recreation centers, libraries,
mental health clinics, and family service agencies;
and announcements in community newspapers. Of
the couples, 98% were legally married and 97% had
at least one child, and the average length of their
relationship was 12 years (range = 1 to 26 years).
Mean ages were 38.53 years for men and 36.12 years
for women. Of the women, 89% were Caucasian,
6.3%
were Asian, 1.5% were East Indian, 1% were
African, 0.5% were First Nations, and 0.5% were
Hispanic. Of the men, 89% were Caucasian, 6.5%
were Asian, 1.9% were East Indian, 1% were His-
panic, 1% were African, and 0.5% were First Na-
tions.
The median education level involved voca-
tional school or some college for both men and
women. The median couple income for the sample
was $50,000 (Canadian) per year. Twenty-seven per-
cent of the couples reported either current or previous
involvement in couple therapy, 6.5% had been in
family therapy, and 12% reported involving their
children in child therapy.
Among these couples, 116 who had an eldest child
7 to 11 years of age also agreed to participate in a
larger study involving both themselves and their chil-
dren. All 116 wives and 79 husbands completed this
part of the study. There were 63 boys (M age = 8.49
years,
SD = 1.71) and 51 girls (M age = 8.80 years,
SD = 1.57) in this subsample.
Procedure
Couples who agreed to participate were mailed a
consent form, a demographic questionnaire, and a
458KERIG
packet of measures. Partners were requested to fill
out their forms independently and were provided
with separate stamped envelopes in which to return
the forms. Parents with school-age children visited a
laboratory at Simon Fraser University, where each
parent and child separately filled out an additional set
of questionnaires. Children completed their measures
in an interview format with a research assistant.
Measures
All couples: CPS. The CPS (Kerig, 1995a) was
designed to measure four dimensions of couple con-
flict (frequency, severity, resolution, and efficacy), as
well as a variety of conflict strategies. Frequency is a
rating of the number of times parents engage in major
and minor conflicts in a year on a 6-point ordinal
scale ranging from once a year or less (scored 1 for
minor conflicts and 2 for major conflicts) to just
about every day (scored 6 for minor conflicts and 12
for major conflicts). Scores for major and minor
conflicts are summed, resulting in possible total
scores ranging from 3 to 18. Severity of problems is
measured by parents' ratings of the degree to which
they disagree about 21 content areas. These content
areas were compiled from the disparate lists con-
tained in the most commonly used marital measures
in the literature, including the DAS (Spanier, 1976),
the Marital Adjustment Test (Locke & Wallace,
1959),
PREPARE (Flowers & Olson, 1986), and the
Marital Agendas Protocol (Notarius & Vanzetti,
1984).
Each content area is rated from 0 (no problem)
to 100 (a severe problem), and the average of these
ratings is used as an overall index of problem sever-
ity. Efficacy is the average proportion of marital
problems partners report they are able to solve (from
0% to 100%), modeled after Notarius and Vanzetti's
(1984) measure. The average percentage of problems
solved is used as the overall measure of efficacy. In
contrast to the Efficacy scale, which asks for a quan-
titative assessment of the percentage of problems
solved, the Resolution scale calls for a more qualita-
tive rating of the emotional tone of the aftermath of
problem-solving attempts. As derived from Rands et
al.'s (1981) scale, ratings are made on 4-point scales
ranging from never (0) to almost always (3) accord-
ing to the frequency with which 13 different resolu-
tions are achieved. These resolutions range from
highly positive and resulting in increased intimacy
(e.g., "We feel closer to one another than before the
fight"; scored 2) to highly negative, involving con-
tinued or escalating acrimony (e.g., "We end up
feeling angry and annoyed with each other"; scored
—2).
The midpoint reflects unclear or partial resolu-
tion, such as abandoning the issue (E. M. Cummings
et al., 1991). These ratings are summed to create the
Resolution scale score, which represents the degree
to which positive or negative affect dominates the
resolution of conflicts.
The CPS also rates a variety of conflict strategies
using a list of 44 tactics derived from the literature on
marital and interpersonal conflict (Christensen &
Heavey, 1990; Gottman, 1995; Noller & White,
1990;
Pruitt & Rubin, 1986; Rausch, 1974; Straus,
1979;
Vuchinich, 1987). Participants rate the fre-
quency with which they and their partners used each
strategy in the previous year; therefore, separate self
and partner conflict strategy subscales can be formed,
or the frequencies can be summed over both scales to
obtain an overall description of the relationship.1
CTS. The CTS (Straus, 1979) is a 17-item
self-
report questionnaire that has been widely used in
studies of marital conflict and aggression. Three sub-
scales are rated: Reasoning (3 items), Verbal Aggres-
sion (6 items), and Violence (8 items). Items are
scored according to how often each tactic has been
used in the previous year by the respondent or his or
her partner; ratings are made on a 7-point scale
ranging from never (0) to more than 20 times (6). The
CTS was validated on a large nationally representa-
tive sample and demonstrated adequate reliability
and internal consistency, with reliability coefficients
ranging from .50 to .69 for the Reasoning scale, from
.77 to .80 for the Verbal Aggression scale, and from
.62 to .83 for the Violence scale.
DAS. Fathers' and mothers' independent ratings
of their marriage were obtained on the DAS (Spanier,
1976),
one of the most widely used measures of the
quality of the marital relationship in the research
literature. This is a 32-item scale on which partners
rate their marital relationship. The validity of the
DAS has been evidenced by its high correlations with
similar marital scales. Acceptable levels of internal
consistency have been established for the subscales.
Only the ratings of overall adjustment (a = .96) were
used in the present study. In the present sample, the
mean overall adjustment scores were 107.44 (SD =
15.30) for fathers and 107.36 (SD = 18.17) for moth-
ers.
Of fathers' and mothers' scores, 24.4% and
21.5%,
respectively, fell within the distressed range
(<97).
Mothers' and fathers' scores were signifi-
cantly correlated (r = .63, p < .001).
Global evaluation of marital quality. Because of
the overlap in content of the CPS and the DAS, high
correlations between the two measures were ex-
pected. As a means of providing a less confounded
index of marital satisfaction with which to assess the
construct validity of the CPS, Fincham and Brad-
bury's (1987) recommendation was followed that
partners provide a global evaluation of their mar-
riage. Partners rated their happiness with their marital
relationship on a 7-point scale ranging from ex-
tremely unhappy (0) to perfect (6).
Satisfaction with conflict
resolution.
Partners also
rated, on a 5-point scale ranging from very dissatis-
1 The complete instrument and scoring instructions
are available from Patricia K. Kerig.
INTERPARENTAL CONFLICT AND CHILD ADJUSTMENT459
fied (0)
to
very satisfied (4), how satisfied they were
with
the
strategies they
had for
conflict resolution.validity, good internal consistency
(K-R 20 = .83),
and concurrent validity with other child anxiety
measures.
Parent
and
Child Measures
Parent reports
of
marital conflict. Mothers
and
fathers rated
the
frequency
of
child exposure
to in-
terparental conflict
on the OPS
(Porter
&
O'Leary,
1980).
This scale consists
of
10 items regarding how
often children are present when their parents argue on
a variety
of
topics
and the
extent
to
which children
are exposed to physical
or
verbal aggression between
their parents. Items are rated
on a
5-point scale rang-
ing from never (0)
to
very often
(4) and are
summed
to create one score
for
exposure
to
interparental con-
flict. The authors have reported good internal consis-
tency
(a = .86) and
reliability
(r = .96)
over
a
2-week period.
Child reports of interparental conflict. Children's
reports
of
their parents' conflicts were obtained
on
the Children's Perceptions
of
Interparental Conflict
questionnaire (CPIC; Grych
et al.,
1992). Children
are presented with
51
statements regarding feelings
they might have when their parents argue (e.g., "I feel
caught
in the
middle when
my
parents have argu-
ments"
or
"When
my
parents argue
I'm
afraid that
something
bad
will happen"),
and
they
are
asked
to
rate whether each statement
is
true
(2),
sort
of
true
(1),
or false
(0). The
CPIC rates children's percep-
tions
on
nine subscales: Intensity
(a =
.82), Resolu-
tion
(a =
.83), Frequency
(a = .70),
Threat
(a =
.82),
Coping Efficacy
(a = .69),
Self-Blame
(a =
.61),
Content
(a =
.74), Triangulation
(a =
.71),
and
Stability
(a =
.65). Factor analysis showed that
the
subscales loaded onto three factors: Conflict Proper-
ties,
Threat, and Self-Blame. The CPIC was validated
on
a
sample
of
222 children
and
cross-validated
on
144 children; good test-retest reliability
and
internal
consistency were established. Children's CPIC
scores correlated with their parents' reports
on the
CTS (Straus, 1979), Marital Adjustment Test (Locke
& Wallace, 1959),
and OPS
(Porter
&
O'Leary,
1980).
The
CPIC also predicted parents' reports
of
children's aggression
and
depression
and
children's
self-reports
of
internalizing symptoms.
Child adjustment. Mothers and fathers rated child
behavior
on the
Child Behavior Checklist (Achen-
bach
&
Edelbrock, 1991),
a
well-validated, reliable,
and widely used measure
of
children's behavior
problems. Factor analyses have confirmed three con-
sistent scales
of
child symptoms: Internalizing (e.g.,
anxiety
and
depression), Externalizing (e.g., aggres-
sion
and
misbehavior),
and
Total Problems.
Child anxiety. Children's ratings were obtained
on
the
Children's Manifest Anxiety Scale-Revised
(Reynolds
&
Richmond, 1978),
a
28-item self-report
questionnaire that assesses
a
child's level
of
manifest
anxiety.
The
instrument
has
demonstrated construct
Results
Factor Analysis
of
the
CPS
Conflict
Strategy Items
The generalized least squares method was
used to enter husbands' and wives' scores on
the 44 conflict strategy items into separate ex-
ploratory factor analyses. It was expected that
the subscales would be related to one another;
therefore, an oblique rotation (oblimin) was
used.
The factor pattern matrices for husbands' and
wives'
scores are shown in Table 1. A six-factor
solution was found to fit the data for both wives,
X2 (551, N = 273) = 522.70, p < .80, and
husbands, x2
(551,
N = 273) = 569.53,p < .28,
and was accepted over other solutions on the
basis of consideration of several criteria, includ-
ing the following: root one (i.e., eigenvalue s
1.00),
scree, and interpretability of the factors
(Johnson & Wichern, 1988). Items were in-
cluded on a factor if the item loading was .30 or
higher; in cases in which an item loaded onto
two factors, the largest factor loading was used
to determine placement. The first factor, termed
Verbal Aggression, included items involving
yelling, accusing, insulting, and so forth. The
second factor involved items reflecting threat-
ening or inflicting harm and was labeled Phys-
ical Aggression. The third factor, labeled Col-
laboration, included items such as trying to
reason with the other person, talking about the
issue, and expressing thoughts and feelings. The
fourth factor assessed nonconstructive strate-
gies and was termed Stalemate, because it
seemed to reflect partners who had reached an
impasse in their attempts to end their quarreling.
The fifth factor, involving such items as trying
to ignore the problem, leaving the scene, and
giving in to the other person to escape argu-
ment, was termed Avoidance-Capitulation. The
final factor, Child Involvement, was formed of
items including arguing in front of the child,
involving the child in the argument, and arguing
about child-related matters.
The items that loaded on the respective fac-
tors were identical for husbands and wives, with
the exception of three items: "Seek intervention
460
£
s
I
"a
a
a
I
E
KERIG
\'
\'
'
i* i"
i
r ( ' ' { '
i
r
§
|
c^ co cs ^H co
oo
oo
^^
co oo ON ^D *~^ ^5 co ^5 *"H
(*••
vi wi co vi 0*4 t*"*
*~* *—*
^5
^f ^p
O *••4 CS
O O O
CS
O O O
O.O
"—ii—*Oi—ION'HCJ
CIO
*—
CS
O O CO O O
«^
r
r '
\' \'
r
" i" i* "
i"'
i"
r r i
r r r '
'
§
vo
to
^*^
*™H
^5
r"**
r^*
c^ ^^
io
C> ^D *O
^f
co
^f
ID
^
v^ ooj <j*
ro
c^J
ON
ooj^f
|
^*|
t*-*!"^!
OHHOH
*-H
OOOOOO
O
i—4
O
i—'
O O O
ro|
O
(N
CS
O
t^ t-;
"^
f^ H;
r
f
' r
*
r
r
i*
*
i" i*
r r r
i*
r ' r
*
*
r i
co
ON
*o
co
io
oo
cococoO'^^t en en
*o
CSIONIOIONI^O
^H|ON|
CNIOICN CS
^"
^H
I^-
^? ^3 ^~< ^^ ^3
^5
i~~( ^3 ^3 ^5 ^5 ^5 ^3 <^ <^> vol^Dj^f
lyDlC^
^Olco '^fl^f
|C*
**H
^3
C3 C*4
i*
r ' r r
r
'
i*
r r '
f
i i r
r
O i-H
o O O '~| O
T"H
t"1-leo
lei jCn
*ol^|^IO cOi-iOrH
O1*
I-HI-HOO
O O
»—<
i
i"
r r r
r
r r
r
|f^-|^O|vo|
nil
ONlenlni vo
HI
o\
C^OCNCNOIVINOV)
©ON
eninv^en
r^
CNCS
en|
rn|en|© ©
cs
© ptNenpppp©
P" pppn © pn
r
'
r '
'
f
'
\~
i
^o ^^ en
t*1^*
^^ ^^ oo ^o co
r*1^"
oo ^o co ^o ON O^ ^^ OQ
i^i
t^j
Q^
f11**
^j
i-^ F-^ r-^
^^
^_^
^5
C2 CN *^ ^? ^i ^i
^^
^5 ^D ^3 ^5
^^
^3 ^5 *~* cs c5 ^5 c*J co
^^
c^J
*•••*
r r r r
r
r
'
r
'
r
r
" '
r
'
r
"
i'
'
r
r r
r
r
^^
oo
*o
^*^ ^3
*o
co ^D
ON
»••<»~* ^p ^5 ^?
*o
""H co ^"^ *"• *~~i
io
»"H
i/~i
co i^f cM
io
*~H *O
'—;
O O O
i-^
»—i OtNOOOO CNOOOO'—inr-1 OO
OOOO
O OO
r
r r ' r
r
r
*
r r r i*
" * i" i* * *
r r r
r r
r r
OO
^^ ^D
CO
f*"1**
^^
CS
^^1
^^
ON ON
^O
00 ^^
^^^
CO
^O
^"1 CO ON
^^ ^D r*"*"
ONICSIOOI ^H I
f^^l^^i
O
O
i-H
O
'~|
*—
COOOOO'"^ HiHfSpOr^pO
*™^
O O
^Kf
1^1
^"1
^'|^|
r
'
i"
r
'
i' ' r
"
i'
r r r r i
\'
i i
i*
i" i* i* i"
' r
' r
i
i "
r
r r '
S
oo
•*£ oo
^j"
»—i
ON
^* Nojinl^lenl HHtenlml©
t
^"*n©
©o>
NQinennn
»n
oo
ON
p
p ©
HI
©
CN HI
oo|r-;|t-;|r-;| r-;|NO|>n|©
© o p p cs©
©ojp©
en pp
I
III' ' I
I*
l'
l"
f I f
tS
-H
©
i
r r
r
"
r r
'
\' \'
en en en en
cs ts
en
f
o H
o\
©
>n
(N en
*
en
-^-
en
INTERPARENTAL CONFLICT AND CHILD ADJUSTMENT461
i
3
S
I
N* OO OO
N ©O© N
83131 31 SPA 31 SI 8 S 53
I' .| . .1 •!•!-
o| co|p p t-;| ©
r
r r ' i
2 2 28 8 S S
CS ON
r—I •—H ©
r r
'
r
ii i i i i i i
ON
© m —oo oo m oo <n
© O N t
CS
OO
ON
©'
m
S 388 S 38 2
r
i i S S 22 H2
i 2;
gN CO ©
p
r
'
r
3
i
SS|
>n —iit—-| o\| r-—ic—II oi en o©
© © ©
oo
r
S 3 S3 8
r
' r SS S©
r
' f
2 ©8S 8 ©2 8 S 8 8 SS B^
r r
S 38S 2
r r r § S S S8 88
r r i i i
2 82§ g tS S : 8 S 82 88
' f
!
'I Hi"3§i 1
a
ill el&
rf*55i«.
!!
M
-9
a Si* a f.
I
o
S
2
II
|I
I
462KERIG
from a counselor/therapist/friend" was associ-
ated with the Stalemate factor for wives and
with the Collaboration factor for husbands;
"Cry" was associated with the Stalemate factor
for wives and did not load on any factor for
husbands; and "Try to smooth things over" was
associated with the Avoidance factor for wives
and with the Collaboration factor for husbands.
Scales were formed by summing unit-weighted
scores only on items that loaded onto their re-
spective factors for both men and women.
The internal consistencies of the conflict
strategy subscales were assessed by computing
coefficient alphas (see Table 2). Scales for both
husbands and wives were found to be reliable.
The intercorrelations among the conflict strat-
egy scales for wives and husbands were calcu-
lated (see Table 3). As expected, there were
strong associations among the factors assessing
noncooperative strategies for negotiating con-
flict, such as Verbal Aggression and Physical
Aggression. There were also negative correla-
tions between scales assessing constructive
strategies (e.g., Collaboration) and nonconstruc-
tive strategies (e.g., Verbal Aggression).
As a means of further assessing the structure
of the CPS, all scales, including the six conflict
strategy scales and the four conflict dimension
scales (Frequency, Severity, Resolution, and Ef-
ficacy), were entered into exploratory factor
analyses performed separately for husbands and
wives.
The factor pattern matrices from both
analyses are presented in Table 4. Although the
solutions did not reach conventional levels of
significance, / (25, N = 267) = 43.79, p <
.01,
for wives and ^ (25, N = 267) = 48.22,
p < .01, for husbands, consideration of criteria
Table 2
Internal Consistencies (Coefficient Alpha) for
Relationship Scales in the Conflicts and
Problem-Solving Scales
ScaleWife Husband
Frequency
Severity
Resolution
Efficacy
Collaboration
Avoidance-Capitulation
Stalemate
Verbal Aggression
Physical Aggression
Child Involvement
.75
.98
.79
.94
.86
.70
.76
.85
.83
.81
.78
.98
.79
.91
.86
.74
.78
.84
.87
.85
00
S
"3
I
!
-S-
1
.3
elations
OS
VO
1
* * *
#
*
*
#
* * *
*
# #
* *
© oo en
Os
en
m m m en en
I
I I
* # # # #
r-1 Tf
eses
en T}- en es
<
'if f
* * * * i
* # # * i
en es <s
Os
in vq vo en Tt
* * # * *
* * * # *
* * * # *
en en
c
en en
es en en
CN
en
' \' \' f
#
# * *
* * # #
*
r r
* * *
*
# #
*
# #
en en
in in in
r
r
i
i
i
i i
tion
#
©
Os t-~
en
VO
* * * #
# *
* # * # * *
# * * * * *
>f- r- O-H
^H es es >n
tion
O I O
.
c 'O
i> 'O
o 0 » 55 o
c
£?-.a
g>g gg s
* *
* *
VO*
es
in
I
I
oo m in vo
en
'-*
en
es
r r
r
*
en
in
f
*
en
m
1
I
1
*
*
*
* *
OseN
en in
* *
^H es
**
en vo
f
*
*
*
vo
1
*
*
*
es in
*
s
1
*
*
*
es
es
in
es
m
*
m
en
*
en
vo
en
#
*
en
vo
oo
es
*
*
# # *
*
*
vo
en
*
*
00
*
*
e*
r
©
es
I
o
1
OS vo
1
en
m
* *
* *
en in
^ m
I
s
18
;££§33u=i£
s^g,,-,
...
Os
1.
3>
V
EMTERPARENTAL CONFLICT AND CHILD ADJUSTMENT463
Table 4
Factor Pattern Matrix From Factor Analysis
of Conflicts and Problem-Solving Scales
WifeHusband
Scale1 1
Frequency
Severity
Resolution
Efficacy
Collaboration
.89 -.26 .63 -.04
JO .17 A9 -.30
-.42 .56 -.39 ,49
.07 49 -.07 41
-.38 .45 .07 .!!_
.06
Avoidance-Capitulation 49 .06 .36 -.09
Stalemate .54 .25 /73 -.16
Verbal Aggression .73 .14 .86 -.07
Physical Aggression 36 .12 J& .12
Child Involvement .50 .15 .77 .01
% of variance
Eigenvalue
Total variance i
30.90 22.60 47.30 11.00
4.72 1.25 4.73 1.10
53.50 58.30
Note. Weights larger than .30 are underscored.
including root one, scree, and interpretability
suggested that a two-factor solution best fit the
data for both husbands and wives, consistent
with two general factors of constructive and
nonconstructive conflict resolution strategies.
Agreement Between Husbands and Wives
on the CPS
Correlations were calculated between part-
ners'
descriptions of their relationship and their
descriptions of one another on the CPS. Hus-
bands and wives agreed in their descriptions of
their relationship, correlations ranging from .26
(p < .001) for Avoidance-Capitulation to .69
(p < .001) for Physical Aggression; the median
correlation was .59. Only correlations for
Avoidance and Efficacy fell below .42. In ad-
dition, correlations were calculated between
husbands' and wives descriptions of the hus-
band's conflict strategies (husband's ratings of
self correlated with wives' ratings of partner)
and between husbands' and wives' ratings of
the wife's conflict strategies (wives' ratings of
self correlated with husbands' ratings of part-
ner).
There was significant agreement between
partners' ratings for all of the subscales. Corre-
lations ranged from .29 (p < .001) for ratings of
wives'
avoidance to .73 (p < .001) for ratings of
husbands' physical aggression, with a median
correlation of .54.
Gender Differences in Conflict Strategies
Gender differences in husbands' and wives'
use of the conflict strategies were assessed by
examining mean differences in partners'
self-
ratings; t tests for paired samples were per-
formed. These analyses showed that husbands
described themselves as higher on Avoidance-
Capitulation than did wives, t(267) = 7.08, p <
.001 (M = 7.44, SD = 3.78, for husbands and
M = 4.79, SD = 2.92, for wives). In contrast,
wives endorsed all of the other conflict strate-
gies more often in their descriptions of them-
selves. The following t values, along with
means for wives and husbands, respectively,
were obtained: Collaboration, t(261) = -2.49,
p < .05, Ms = 15.47 (SD = 2.50) and 15.00
(SD = 2.43); Stalemate, t(261) = -12.15, p <
.001,
Ms = 7.44 (SD = 3.78) and 4.70 (SD =
2.92); Verbal Aggression, t(261) =
-5.05,
p <
.001,
Ms = 12.41 (SD = 4.61) and 11.00
(SD = 4.53); Physical Aggression f(267) =
-3.15,
p < .01, Ms = 1.54 (SD = 2.25) and
1.84 (SD =
1.12);
and Child Involvement,
t(267) = -2.44, p < .05, Ms = 5.89 (SD =
3.16) and 5.39 (SD = 3.05).
Stability
Forty-eight couples filled out the CPS a sec-
ond time, approximately 3 months after they
first completed the measure. Test-retest corre-
lations between the two times were acceptably
high for each of the scales, ranging from a low
of .53 (p < .01) for Child Involvement to a high
of .87 (p < .001) for Severity, with a median
correlation of .63.
Convergent and Discriminant Validity
Correlations between wives' and husbands'
ratings on the CPS and the DAS, CTS, and OPS
are displayed in Table 5. Significant correla-
tions were found between overlapping scales of
the CPS and other measures of marital conflict.
For instance, there were positive correlations
between the CTS Reasoning scale and the CPS
Collaboration scale, as well as between the Ver-
bal Aggression and Physical Aggression scales
on the two measures. Divergent validity was
indicated by the nonsignificant correlations be-
tween dissimilar scales of the two measures.
Similarly, the CPS scales and the subscales of
464
KERIG
e
•a
2
I
•r
60
I I I
* * * *
* *
# #
*
* * * * *
*
en
I I I
*
#
O
*
Ti
ON
«
* * * *
o -<t
g;
O
# #
*
# #
I I I
*
CS
©
I
00
§5
00
Tf
r-
\o
CS
00
irt
I I
* *
#
\O
-H
tS
en vo vo
8SS
* #
#**
en o> cs
en*\o
O\ O\
so
en
r r
*
II V
INTERPARENTAL CONFLICT AND CHILD ADJUSTMENT465
the DAS showed high correspondence. Corre-
lations between parents' OPS ratings of child
exposure to interparental conflict and the CPS
Child Involvement scale were also significant
for both husbands and wives.
Prediction of Global Ratings of
Marital Happiness
As a means of determining whether the CPS
scales were predictive of marital quality, multi-
ple regressions were performed with the global
ratings of marital happiness as the dependent
measure and CPS subscales as predictors, en-
tered simultaneously. For husbands, the CPS
predicted a significant proportion of variance in
global marital quality, R2 = .54,