Content uploaded by Anne Martin
Author content
All content in this area was uploaded by Anne Martin on Feb 08, 2019
Content may be subject to copyright.
Running head: MOTHER AND FATHER PARENTING
The Joint Influence of Mother and Father Parenting
on Child Cognitive Outcomes at Age 5
Anne Martin, Rebecca M. Ryan and Jeanne Brooks-Gunn
National Center for Children and Families
Teachers College, Columbia University
Revised June 9, 2007
Address correspondence to:
Anne Martin
Teachers College, Columbia University
525 West 120th Street, Box 39
New York, NY 10027
TEL: 212-678-3479
FAX: 212-678-3676
EMAIL: arm53@columbia.edu
Mother and Father Parenting 1
Abstract
Few studies of parenting have considered the possibility that the association between one
parent’s supportive parenting and a child’s early cognition is moderated by the other parent's
supportiveness. We test this proposition using a low-income sample of coresident couples. In
addition, we cross-classify parents within couples according to their parenting behaviors to test
for homogamy. Mothers and fathers were videotaped during separate free-play dyadic
interactions with their two-year-old child, and each parent's behavior was coded according to six
scales. K-means cluster analysis was used to describe each parent’s pattern of parenting
behaviors. Parents were then cross-classified within couples by pattern. There was mixed
evidence of homogamous parenting styles within couples. After parents were consolidated into
four primary pairings, children in each of these pairings were compared on age five math and
language scores. Children with two supportive parents scored highest, while those with two
unsupportive parents scored lowest. Among children with one supportive parent, the sex of that
parent was inconsequential. There were no significant interactions between maternal and paternal
supportiveness on either math or language. Thus it appears that in this low-income sample the
combined effects of maternal and paternal supportiveness are additive.
Mother and Father Parenting 2
The Joint Influence of Mother and Father Parenting on Child Cognitive Outcomes at Age 5
The overwhelming majority of studies on the influence of parenting practices on young
children’s cognitive development have focused exclusively on mothers. A substantive body of
literature now demonstrates the link between maternal supportiveness – defined variously as
acceptance, warmth, responsiveness (or contingency), and stimulation – and positive early child
outcomes (e.g., Cowan, Cowan, Heming, & Miller, 1991; Landry, Smith, Swank, Assel, &
Vellet, 2001; NICHD Early Child Care Research Network, 1999; Pettit, Bates, & Dodge, 1997;
Tamis-LeMonda, Bornstein, & Baumwell, 2001). Although fewer studies have focused on
fathers’ influence on young children’s cognitive development, those that have been conducted
have found that as with mothers, responsiveness, sensitivity, and stimulation from fathers predict
superior outcomes (e.g., Black, Dubowitz, & Starr, 1999; Kelley, Smith, Green, Berndt, &
Rogers, 1998; Lamb, 1997; NICHD Early Child Care Research Network, 2004; Tamis-
LeMonda, Shannon, Cabrera, & Lamb, 2004).
Despite increasing interest in fathers’ parenting, surprisingly little attention has been paid
to the combined effects of mother and father parenting practices. The studies establishing the
relevance of father parenting to child outcomes have typically done so without considering the
mother’s concurrent influence (e.g., Black et al., 1999; Kelley et al., 1998) or have adjusted for
that influence in order to isolate the independent contribution of fathers (e.g., NICHD Early
Child Care Research Network, 2004; Tamis-LeMonda et al., 2004). However, children in two-
parent families experience a single home environment that blends both parents’ influences. Both
life course theory (Elder, 1998) and family systems theory (Minuchin, 1985) have enriched our
understanding of human development by emphasizing the interdependence of family members.
An analysis of parental influence that accounts for both parents’ behavior should more closely
Mother and Father Parenting 3
approximate children’s lived experiences than one that attempts to measure each parent’s
influence net of the other. Therefore, one advantage to the concurrent study of both parents is
that it should allow for a richer understanding of young children’s developmental context.
A second advantage to studying parents in combination is that it may be possible to
identify interactions between the effects of mothers’ and fathers’ parenting that are not visible in
studies of main effects. For example, based on the literature linking maternal and paternal
supportiveness to early child cognition, it might be expected that the effects of a supportive
mother and a supportive father are additive. However, there may be a synergistic dynamic
between two supportive parents that produces a multiplicative effect. In this scenario, the
increase in outcome score associated with two supportive parents would exceed the sum of the
increases associated with each parent separately. Parents who are supportive may have more
refined interpersonal skills than other parents and hence have higher-quality relationships. They
may also have fewer disagreements over child-rearing and thus less marital conflict. Brody,
Pillegrini, and Sigel (1986) found that couples with fewer marital problems had more similar
parenting behaviors during teaching tasks than couples with greater marital problems. Although
the direction of this association is indiscernible, it suggests that parents with concordant styles
may have greater relationship harmony, which in turn predicts more optimal parenting practices
(for a review see Krishnakumar & Buehler, 2000), better parent-child relationships (for a review
see Erel & Burman, 1995), and higher cognitive test scores in young children (van Bakel &
Riksen-Walraven, 2002).
Alternatively, non-linear effects of supportiveness may operate in the opposite direction,
such that the increase in outcome score associated with two supportive parents is smaller than the
sum of the increases associated with each individually. In this case, a supportive parent would
Mother and Father Parenting 4
confer a smaller advantage than expected when the other parent was also supportive. This
scenario might arise if there were a ceiling on potential increases in the outcome. Yet a third
possibility is that the increase in outcome score associated with a supportive parent added to an
unsupportive parent is larger than the sum of the increases associated with each individually. In
this scenario, a supportive parent would be particularly beneficial in the presence of a non-
supportive one. Finally, it is possible that the increase associated with a supportive parent is
diminished when the other parent is not supportive because a non-supportive parent undermines
or even offsets the benefits conferred by the supportive parent. For example, a child with a
highly supportive parent whose other parent is extremely harsh and punitive may be unable to
fully benefit from the first parent’s support. Indeed, the extent to which a child benefits from a
single supportive parent may depend on the particular style of the unsupportive parent, with
some negative practices serving as stronger counterweights than others.
In short, the benefits to the child of a supportive parent may be moderated – by being
either enhanced or eroded – by the second parent’s parenting behaviors. The likelihood that
parents’ parenting behaviors interact is suggested by several studies showing that the influence
of a parent’s mental health turns on the mental health of the other parent. These studies find that
the children of mothers with a mental health problem are at far greater risk of exhibiting behavior
problems when their father also has a mental health problem (Dierker, Merikangas, & Szatmari,
1999; Goodman, Brogan, Lynch, & Fielding, 1993; Kahn, Brandt, & Whitaker, 2004). Given the
close association between mental health and parenting (for reviews see Goodman & Gotlib,
1999; Lovejoy, Graczyk, O’Hare, & Neuman, 2000), it is likely that similar contingencies apply
to parenting style, and that each mother-father combination has a unique influence on children.
Mother and Father Parenting 5
One of the studies on the combined effects of maternal and paternal mental health found
that when only one parent was highly distressed, the child developed more behavior problems if
that parent was the mother (Kahn et al., 2004). This finding points to the potential relevance of
parent sex to the investigation of parental supportiveness. To our knowledge, no studies to date
have examined whether the combination of a supportive mother and unsupportive father exerts a
different influence on young children’s cognition than the combination of a supportive father and
unsupportive mother. The literature suggests divergent hypotheses. The benefits of a supportive
mother may outweigh those of a supportive father because mothers generally play a more central
caregiving role (Pleck, 1997). However, mothers’ and fathers’ parenting practices tend to have
more similar than dissimilar influences on child outcomes (Lamb, 1997). Therefore, there may
be no differences between the effects of a supportive mother and a supportive father.
The question of whether one parent’s parenting behavior pattern moderates the other’s
takes on extra significance if there are certain pairings of parenting patterns that occur with
greater frequency than others among couples. To date, there is no literature on the combination
of preschool children’s mothers and fathers with respect to observed parenting practices. With
one exception (Belsky & Fearon, 2004), none of the studies with observational data on parenting
practices for both parents of young children has cross-classified parents within couples (e.g.,
Belsky, Gilstrap, & Rovine, 1984; Brachfield-Child, 1986; Brody et al., 1986; Conner, Knight, &
Cross, 1997; Kerig, Cowan, & Cowan, 1993; Power & Parke, 1983; Stevenson, Leavitt,
Thompson, & Roach, 1988; Tamis-LeMonda et al., 2004). Gauvain, Fagot, Leve, and Kavanagh
(2002) compared five-year-old children's performance on a parent-child instructional task
according to whether neither, one, or both parents scored high on task information, but did not
report the size of these groups.
Mother and Father Parenting 6
It is reasonable to expect similarity within couples, both because of assortative mating
and because of mutual influence. Indeed, Belsky and Fearon (2004) found that mothers’ and
fathers’ scores on sensitivity toward their child in the first four years of life were moderately
correlated. On the other hand, one might expect sex differences to outweigh similarities if
gendered parenting practices tend to recur across couples (Lovas, 2005). In addition, couples
may adopt different but complementary parenting behaviors to capitalize on each parent’s
strengths and compensate for each one’s weaknesses.
The present study assesses the association between maternal and paternal supportiveness
during toddlerhood and young children’s subsequent cognitive development in low-income two-
parent homes. The advantage of the dataset selected for this analysis is that observational
measures of parenting are available for both mothers and fathers, based on mother-child and
father-child interactions videotaped when children were approximately two years old. Practical
constraints did not allow for the mother, father and child to be videotaped together in a single
interaction. While observations of triadic interactions would have been optimal, the available
parenting measures based on dyadic interactions are still meaningful, given that children in two-
parent families typically spend some time alone with each parent.
Our dependent variables are measures of math and language achievement at age five.
This study extends an earlier analysis which found associations between maternal and paternal
supportiveness, as measured by the videotaped interactions when children were two, and
children’s scores on the Bayley Mental Development Index (MDI; Bayley, 1993) at two and
three years of age (Ryan, Martin, & Brooks-Gunn, 2006). We use the age two classifications of
mother and father parenting to enable comparisons of findings at age five with those at ages two
and three. These classifications characterize mothers and fathers by their membership in groups
Mother and Father Parenting 7
generated by cluster analysis of parenting behavior scales from the dyadic interactions.
Associations between cluster membership and child MDI were evident both when each parent
was analyzed separately and when parents were analyzed jointly as a couple according to the
cross-classification of their clusters. The present study extends this analysis with the next wave
of data by testing for sustained associations between parental supportiveness and children’s math
and language scores at age five. First, we examine associations between cluster membership and
child outcomes separately for mothers and fathers. Next, mothers and fathers are cross-classified
within pairs. These cross-classifications are then used to predict child outcomes, net of family
background characteristics and the child’s earlier outcome score.
This study is a useful complement to our last study because supportive parenting during
early childhood appears to have a cumulative beneficial effect over time. Landry et al. (2001)
found that children whose mothers were supportive in the first four years of life demonstrated
higher cognitive scores than children whose mothers were supportive in the first two years alone.
It is possible that in our sample the gap between children whose parents were supportive at age
two and those whose parents were not may be larger at age five than it was at ages two and three.
In addition, by examining child outcomes three years after parenting was measured, while
controlling for outcomes measured at the same time as parenting, this study strengthens our
confidence that what we are labeling effects of parenting behaviors on child cognition are not
actually attributable to reverse causality (child cognition causing parenting behaviors) or
confounding (a third variable causing both).
Our earlier paper subjected maternal and paternal parenting behavior scales to separate
cluster analyses to determine if the behaviors covaried differently by parent sex (Ryan et al.,
2006), which research on differences in maternal and paternal parenting styles might suggest
Mother and Father Parenting 8
(Brachfeld-Child, 1986; MacDonald & Parke, 1984). We found, in fact, that they did not, and
that similar classification schemes could be used to describe parenting behaviors in mothers and
fathers. In this paper, we use these separate classifications to report on maternal and paternal
pairings within couples, with special attention paid to the question of whether partners are
homogamous with respect to parenting behaviors.
Finally, there is particular value in describing the children’s cognitive status at age five,
given their impending entry into school. Longitudinal data have demonstrated continuity
between academic performance in pre-kindergarten, kindergarten, or first grade and later
academic performance in elementary school (Duncan, Claessens, & Engel, 2005; Shaywitz et al.,
1995; Williamson, Appelbaum, & Epanchin, 1991) and high school (Ensminger & Slusarcick,
1992; Stevenson & Newman, 1986).
Our research strategy addresses four research questions: (a) What are the separate
associations between mothers’ and fathers’ parenting patterns at age two and their children’s
cognitive outcomes at age five? Does age two parenting predict age five cognitive outcomes
even when age two cognitive outcomes are accounted for? (b) What configurations are produced
by cross-classifying mothers and fathers within couples according to parenting pattern? Are
couples generally concordant or discordant with respect to their pattern of parenting behaviors,
and are certain combinations of parenting patterns more prevalent than others? (c) What are the
associations between mother-father parenting combinations and young children’s cognitive
outcomes? In particular, do children with two supportive parents score significantly higher than
children with only one? Do children with one supportive parent score significantly higher than
children with none? Are the benefits of two supportive parents additive or multiplicative? Do the
Mother and Father Parenting 9
results vary by developmental domain (language vs. math)? (d) Among children with one
supportive parent, does it matter whether it is the mother or father?
Our hypotheses are guided by our own and others’ previous research. With respect to the
first research question, it is expected that children with a supportive mother at age two will score
higher than children without one on math and language tests at age five, and that the same will
be true for children with a supportive father. We previously found that, adjusted for background
characteristics, at ages two and three, children with a supportive mother scored higher on the
MDI than other children, and children with a supportive father scored higher than those without
one, although statistically significant contrasts obtained only for the children of the most highly
supportive parents (Ryan et al., 2006). Among both mothers and fathers, the spread between the
highest and lowest scoring groups widened between ages two and three. Consequently we expect
that by age five these differences will widen further, such that the advantage experienced by
children with a supportive parent will be pronounced.
Regarding our second research question, there is no empirical basis for the prediction of
cross-classified frequencies of mother and father parenting patterns owing to a lack of evidence
to date. However, in the NICHD Study of Early Child Care, mothers’ and fathers’ scores on
sensitivity from separate dyadic interactions with the child during the first four years of life have
been fairly to moderately correlated within couples (Belsky & Fearon, 2004; NICHD Early Child
Care Research Network, 2004). Additional studies have found moderate-to-high correlations
within pairs on self-reported parenting scales (Bentley & Fox, 1991; Hart, DeWolf, Wozniak, &
Burts, 1992; Winsler, Madigan, & Aquilino, 2005). Further, the literature on within-couple
homogamy demonstrates that men and women select marriage and cohabitation partners of
similar race, education, and religion (Blackwell & Lichter, 2000; Kalmijn, 1998; Qian, 1997;
Mother and Father Parenting 10
Schoen & Weinick, 1993). Population studies also suggest a high degree of assortative mating
for psychosocial characteristics such as antisocial behavior and depression (Galbaud du Fort,
Bland, Newman, & Boothroyd, 1998; Krueger, Moffitt, Caspi, Bleske, & Silva, 1998). Because
demographic and psychological characteristics are highly associated with parenting behaviors
(McLeod & Nonnemaker, 2000; McLoyd, 1990, 1998; McLoyd, Jayaratne, Ceballo, & Borquez,
1994), there is reason to suspect that couples also exhibit similar parenting patterns.
Expectations about the associations between mother-father supportiveness combinations
and children’s math and language scores at age five are based on our earlier findings. At ages
two and three, children with two supportive parents scored higher than all other children, while
children with two unsupportive parents scored lowest (Ryan et al., 2006). At both timepoints
children with two supportive parents scored higher than children with one, although this
difference was only significant at age three. Children with one supportive parent scored higher,
though not significantly so, than those with none at both timepoints. We expect children with one
supportive parent to continue to score higher than those with no supportive parents at age five.
We also expect children with two supportive parents to continue to score higher than those with
one.
It is difficult to predict whether the combined effects of two supportive parents on child
cognition will be additive or multiplicative, given the absence of previous tests of this question.
The literature on parental mental health, however, suggests that maternal and paternal
supportiveness may have a negative multiplicative effect. Past findings that children with one
mentally healthy parent are behaviorally comparable to those with two mentally healthy parents
indicate that parental mental health may have a threshold effect on children. Because parents’
mental health and parenting behaviors are closely associated (Goodman & Gotlib, 1999; Lovejoy
Mother and Father Parenting 11
et al., 2000), supportive parenting may also have a threshold effect on children. In such a
scenario, once a certain level of supportive parenting is met, there would be diminishing returns
to the child. In point of fact, the suggestion of diminishing returns was present in our sample at
age two, when the first supportive parent was associated with a greater increase in MDI score
than was the second supportive parent (although this differential did not achieve statistical
significance; Ryan et al., 2006). Therefore, it is reasonable to expect that at age five, the
advantage experienced by children with two supportive parents may be smaller than expected,
based on the increase in outcome associated with each supportive parent individually.
With respect to the question about whether the sex of a sole supportive parent matters,
our earlier study found that children with one supportive parent scored similarly at ages two and
three regardless of whether that parent was the mother or father (Ryan et al., 2006). Therefore we
expect similar findings when the children reach five years of age.
Method
Sample
The data are drawn from the Early Head Start Research and Evaluation Project, an
ongoing evaluation of 17 Early Head Start (EHS) programs nationwide. EHS is a federally
funded program for low-income infants and toddlers and their families designed to improve
young children’s school readiness and their caregivers’ parenting skills. All families seeking
enrollment at the programs selected for the evaluation between February 1996 and October 1998
were eligible for inclusion in the study. Families were randomly assigned to treatment and
control groups; the former were offered EHS services and the latter were not.
Children were visited in their homes by trained assessors when they were 14, 24, 36 and
60 months old. At each timepoint, videotaped observations were made of mother-child
Mother and Father Parenting 12
interactions. The present study analyzes video data from the 24-month observation, along with
data from a substudy conducted by eight of the sites which also collected videotape data on
fathers. During the 24-month home visit, mothers were asked permission to contact the child’s
father. Fathers who agreed to participate in the study were videotaped in a dyadic interaction
with their child.
All fathers interviewed at 24-months in the eight sites participating in the substudy who
reported coresiding with the child’s mother were eligible for inclusion in the present study (n =
343). The father’s report of residency status was used instead of the mother’s because fathers
were less likely to report coresidence; therefore, their reports were more conservative. However,
the mother’s report of coresidence was used to test for selection bias in the sample of fathers who
were interviewed at 24 months because we lacked father report of residency for men who were
not interviewed. Of the 665 fathers whom mothers reported as coresident, 322 (49%) were not
interviewed at 24 months. Fathers with interview data (n = 343) and fathers without (n = 322)
were compared on maternal race, age, education level, poverty level, and employment status, as
well as child sex and birth order. The samples only differed on maternal race; mothers with
father interview data were significantly more likely to be White or Black and less likely to be
Hispanic or other race (p < .001, χ2 = 20.23).
Of the 343 mothers and fathers interviewed at 24 months, only 237 had video data for
both parents. Of those couples, 200 had at least one child cognitive outcome score at age five.
These cases constituted our final sample. They were compared to the excluded 143 cases on the
maternal and child characteristics listed above, as well as paternal education, race, age, and child
cognitive scores. There were no significant differences on any maternal or paternal demographic
Mother and Father Parenting 13
characteristics. Children included in our sample had slightly higher Woodcock-Johnson Applied
Problems scores (p < .10, F = 3.7).
Because our sample was drawn from applicants to the EHS program, which targets low-
income families, it was economically disadvantaged. The vast majority (82%) lived below the
federal poverty threshold at baseline (before treatment group assignment). Although the sample
was economically homogeneous, it was ethnically diverse. Sixty-six percent of the mothers were
White, 19% were African-American, 13% were Hispanic, and 3% were of other race/ethnicity.
Sixty percent of the fathers were White, 19% were African-American, 16% were Hispanic, and
5% were of other race/ethnicity. Two-thirds (67%) of mothers and three-fourths (75%) of fathers
had a high school degree or GED at baseline. None of these characteristics was associated with
program status (intervention vs. control group).
Although the term parents is used throughout this article, in rare cases there were
alternative caregivers. Six of the mothers were actually grandmothers (n = 4), an aunt (n = 1), or
a foster mother (n = 1). Fourteen of the fathers were in fact grandfathers (n = 13) or an uncle (n =
1). The remaining children were living with their biological father (81%), adoptive father (9%),
or father figure (4%).
The mean age of the children was 25 months (Range: 23 – 33) during the mother-child
interaction visit and 27 months (Range: 23 – 33) during the father-child interaction visit.
Children’s mean age at the five-year visit was 62 months (Range: 50 – 74). Half the children
(48%) were male, and 55% were first-borns. Exactly half of the children were assigned to the
intervention group and thus received EHS services. All analyses controlled for program status.
Procedures
Mother and Father Parenting 14
Maternal and paternal parenting behaviors were assessed at separate home visits
scheduled as close to the child’s second birthday as possible. The protocol used to record parent-
child interactions was adapted from Vandell (1979) and the NICHD Study of Early Child Care’s
(NICHD Early Child Care Research Network, 1999) “Three Box” free play assessment. At each
visit, a trained home visitor videotaped the parent and child playing with toys contained in three
pillow cases. For mothers, these toys included a book, a set of pots and pans, and a Noah's ark
and animals set. For fathers, these toys included a book, a pizza delivery set of toys, and a farm
set. (For details on the protocol, see Love et al., 2002.) Parents were given intentionally vague
instructions so as to elicit naturally occurring parenting behaviors.
Both the parent’s and child’s behavior were coded; only the parent’s behavior is
considered here. The coding scheme (Brady-Smith, O’Brien, Berlin, & Ware, 1999), adapted
from the NICHD Study of Early Child Care’s “Three Box” coding scheme (NICHD Early Child
Care Research Network, 1999), was guided by theory about the importance of attachment
(Ainsworth, Blehar, Waters, & Wall, 1978; Bowlby, 1982) and scaffolding (Berk & Winsler,
1995; Pratt, Kerig, Cowan, & Cowan, 1988; Rogoff, Mistry, Goncu, & Mosier, 1993; Vygotsky,
1978) to children’s early cognitive development. It included the following six scales of parent
behavior: sensitivity (taking the child’s perspective, accurate perception of the child’s signals,
and prompt and appropriate responses to these signals), positive regard (demonstration of love,
respect, and admiration), cognitive stimulation (teaching or actively trying to expand the child’s
abilities), detachment (lack of attention to and awareness of the child’s cues and absence of
engagement in the child), negative regard (hostility towards or rejection of the child) and
intrusiveness (physical or verbal overcontrol of the child’s play). All scales had seven points,
Mother and Father Parenting 15
ranging from very low (one) to very high (seven). Intercorrelations among the scales are
presented in Table 1.
Videotapes were scored by the National Center for Children and Families (NCCF) at
Teachers College, Columbia University. A research scientist trained a team of coders and was
the “gold standard” for all reliability tests. Coders were trained using sample interactions
selected to exemplify high, medium, and low scores on each scale for mothers and fathers. The
coders achieved reliability (exact agreement within one point) to a criterion of 85% with the
researcher, after which 15% of tapes were drawn randomly from the coders’ weekly assignment
and checked for interrater reliability (again at 85%). Coders’ average agreement for parenting
scales ranged from 89% to 98% for mother tapes and from 94% to 96% for father tapes. All
coders were unaware of the treatment status and any other identifying information about the
videotaped families.
As close as possible to the time of the child’s fifth birthday, interviewers visited the home
and conducted a parent interview, videotaped the mother and child engaging in semi-structured
tasks, and administered a range of child cognitive and socioemotional assessments. The child
cognitive outcomes at age five examined in the present analysis are drawn from this home visit.
Parents were given $20 and a small gift for the child after the 2-year and 5-year visits.
Measures
Mother and father parenting clusters. Parenting clusters were generated separately for
mothers and fathers based on their scores on the parenting scales from the videotaped dyadic
interactions. A non-hierarchical method of cluster analysis such as K-means was preferable to a
hierarchical method because the clusters were assumed to be discrete rather than nested groups
(Henry, Tolan, & Gorman-Smith, 2005). To theorize that the groups were nested would have
Mother and Father Parenting 16
assumed that parents' scores on the six scales fell along a single continuum of supportiveness. In
point of fact, the parenting literature recognizes multiple dimensions of supportiveness, which
has given rise to the use of typologies to describe combinations of parenting behaviors
(Baumrind, 1967; Darling & Steinberg, 1993; Maccoby & Martin 1983; Steinberg, Lamborn,
Darling, Mounts, & Dornbusch, 1994).
In accordance with Aldenderfer and Blashfield (1984), the criteria used to select the
number of clusters included their interpretability, size, and ability to differentiate the outcome
measures. Solutions with two, three, and four groups were attempted for both mothers and
fathers. The four-group solution was ultimately selected for both parents because it yielded the
clearest distinctions among the within-cluster patterns of parenting scales and thus maximized
our ability to characterize each cluster’s pattern of behaviors. For example, although the three-
group solution yielded Unsupportive-Negative and Unsupportive-Detached clusters for mothers,
their means on the scales of negative regard and detachment, respectively, were lower than the
smaller and more selective clusters of the same names produced by the four-group solution.
Mothers and fathers generated similar parenting pattern clusters (Table 2). Because the
clusters were comparable across parent sex, they were given the same names. Clusters were
described according to the pattern of their scale means relative to one another, as well as relative
to those of other clusters.’ In other words, the highest and lowest scoring scales both within and
across clusters were of interest.
The first cluster was called Highly Supportive because it had the highest mean scores on
the positive scales (sensitivity, positive regard, and cognitive stimulation) and the lowest mean
scores on the negative scales (negative regard, intrusiveness, and detachment) among both
parents. The second cluster, called Somewhat Supportive, scored next highest on the positive
Mother and Father Parenting 17
scales and next lowest on the negative ones among both parents. For both sexes, the positive
scale scores were higher than the negative scale scores. The third cluster, called Unsupportive-
Negative, scored higher than the other three clusters on both negative regard and intrusiveness
among both parents, and the fourth, called Unsupportive-Detached, scored highest on
detachment among both parents. Fathers in the Unsupportive-Negative cluster scored higher on
intrusiveness than any of the positive scales, while mothers in the Unsupportive-Negative cluster
scored higher on intrusiveness than positive regard alone. Mothers in the Unsupportive-Detached
cluster scored higher on detachment than any of the other scales except sensitivity, while
Unsupportive-Detached fathers scored higher on sensitivity and cognitive stimulation than
detachment.
Although mothers’ and fathers’ clusters resembled each other, they did not share the
same distribution. Of the 200 families, the clusters in order of size among mothers were Highly
Supportive (41%), Somewhat Supportive (35%), Unsupportive-Negative (15%), and
Unsupportive-Detached (10%). Their order among fathers was Somewhat Supportive (42%),
Highly Supportive (34%), Unsupportive-Detached (15%), and Unsupportive-Negative (9%).
Children’s cognitive outcomes. Children’s math achievement at age five was assessed by
trained assessors with the Woodcock-Johnson-Revised Applied Problems subtest (Woodcock &
Johnson, 1989). Children’s receptive language ability was assessed at age five with the Peabody
Picture Vocabulary Test-III (Dunn & Dunn, 1997). Both measures are age-standardized (M =
100, SD = 15). In the present sample, the mean of the Applied Problems subtest was 92.6 (SD =
18.7), and the mean of the PPVT-III was 95.0 (SD = 16.9).
Covariates. Demographic variables were added as covariates to all models of child
outcomes to increase the precision of estimated parenting effects, after sensitivity analyses (not
Mother and Father Parenting 18
shown) revealed that their inclusion did not substantively alter the study’s conclusions.
Demographic information was collected from mothers at baseline and from fathers at the two-
year visit. Maternal race was coded as non-Hispanic White, non-Hispanic African-American, and
Hispanic. Highest parental education was a trichotomous variable indicating whether at least one
parent had more than a high school degree or GED, at least one parent had a high school degree
or GED, or neither parent had a high school degree or GED. Paternal biological status is an
indicator variable denoting whether the father was the child’s biological father. Poverty status is
an indicator variable denoting whether the mother was above the poverty threshold at baseline.
Program status is coded 1 if the child was in the EHS intervention group.
Because of the likelihood that the child’s cognitive status at age two covaried with both
parenting at age two (our independent variable) and cognitive status at age five (our dependent
variable), it was included as a control in all models. It was not possible to include scores on the
precise math and language assessments from age 5 because they were not administered at age
two. Instead, at age two children were assessed with the MDI (Bayley, 1993), a broad measure of
cognitive ability (M = 100, SD = 15). In our sample, the mean MDI score was 91.8 (SD = 14.2).
The correlation between the MDI at age 2 and the Woodcock-Johnson at age 5 was .53 (p < .01),
and the correlation between the MDI at age 2 and the Peabody Picture Vocabulary Test – III at
age five was .60 (p < .01).
Models were also run with mother and father parenting at age five as covariates so that
the effects of contemporaneous parenting on age five outcomes was not mistakenly attributed to
age two parenting. The protocol used to observe the parent-child interactions at age five differed
from the one at age two in that instead of using three bags of toys, it relied on a single activity
involving Play-doh, a rolling pin, and a cookie cutter. As at age two, mothers and fathers were
Mother and Father Parenting 19
videotaped separately with their child. The coding scheme was revised at age five, such that the
sensitivity and positive regard scales were replaced with a single scale called supportiveness. The
process used to generate clusters based on the age two parenting data was repeated with the age
five parenting data. Clusters similar to those from age two emerged.
The study findings did not change substantively when age five parenting clusters were
added to models predicting child cognitive outcomes. Moreover, age five parenting itself did not
predict child outcomes once age two parenting was accounted for. The most likely explanation
for the absence of a cross-sectional association between child outcomes and parenting at age five
once age two parenting was accounted for is measurement error in the parenting scales at age
five. The Play-doh activity may not have lent itself to joint participation and reciprocity as much
as the age two activity, which used toys requiring adult assistance. It also involved one toy
instead of three, and therefore may have elicited a smaller range of parental behaviors. Finally,
the consolidation of two scales into one may have further constrained potential variability in
parenting behaviors. Collinearity between age two and five parenting is not a likely explanation
for the lack of predictive value of age five parenting, given the lack of comparability in both the
task and the scales. Cluster assignments at ages two and five showed a degree of consistency that
was greater than chance but still only moderate (r = .21, p < .05).
Thus the results reported here are based on a model without age five parenting.
Additional covariates (mother is other race, child’s sex, either parent was a teenager at the
child’s birth, child’s birth order, father’s race) were initially included in models but later dropped
due to non-significance.
Results
Associations Between Mother and Father Parenting and Child Cognitive Outcomes
Mother and Father Parenting 20
Our first research question, which asks whether maternal and paternal parenting patterns
are individually associated with children’s cognitive outcomes, was addressed by using mother
and father clusters to predict child math and language scores at age five. Estimated marginal
means for each mother parenting cluster, and then for each father parenting cluster, were
calculated with maternal race, highest parental education level, father’s biological status,
program status, maternal poverty status, and age two MDI score set to their means in all
equations.
On both math and language, children with a Highly Supportive mother scored highest and
children with an Unsupportive-Detached mother scored lowest, net of maternal and child
covariates (Table 3). The differences between these two groups were sizable. On math, children
with a Highly Supportive mother scored 9.8 points, or 65% of a standard deviation, higher than
children with an Unsupportive-Detached mother (p < .05). On language, they scored 8.6 points
higher (57% of a standard deviation, p < .05).
Children with a Highly Supportive mother scored higher than children with a Somewhat
Supportive mother, but not with statistical significance. Similarly, children with an
Unsupportive-Negative mother scored higher than children with an Unsupportive-Detached
mother, but not with statistical significance. Thus it appeared that the most salient distinction
among mother clusters was between the two Supportive groups (Highly and Somewhat
Supportive) and the two Unsupportive groups (Negative and Detached).
Similar findings emerged for father clusters (Table 3). Children with a Highly Supportive
father scored highest on math and language. Among mothers, the Unsupportive-Detached cluster
had scored lowest, but among fathers, the Unsupportive-Negative cluster scored lowest.
However, children with Unsupportive-Negative and Unsupportive-Detached fathers did not
Mother and Father Parenting 21
differ with statistical significance from each other. The span between children with a Highly
Supportive father and those with an Unsupportive-Negative one was 10.7 points, or 71% of a
standard deviation, on math (p < .05), and 7.4 points, or 49% of a standard deviation, on
language. As with mothers, children with a Somewhat Supportive father did not differ
significantly from those with a Highly Supportive one. Thus, as with mothers, the clearest
distinction appeared between children with a Supportive (Highly or Somewhat) father and those
with an Unsupportive (Negative or Detached) father. Also, as with mothers, results were more
pronounced for math than language.
Mother and Father Parenting Pairs
To answer our second research question regarding parenting homogamy we cross-
classified couples according to the mother’s and father’s cluster assignments. A visual inspection
of the table reveals whether particular combinations of parenting patterns were more common
than others. A chi-square statistic was used to test formally for the presence of homogamy.
Table 4 presents the cross-classification; each cell contains the observed count, expected
count, row percentage and column percentage. A cell's expected count is based on the joint
probability of having a mother and father in those two clusters given the size of those clusters
within the sample. For example, given that 41% of mothers and 34% of fathers are Highly
Supportive, due to chance alone we should expect to find 28 couples in this sample who are both
Highly Supportive (.41 * .34 * 200 = 27.9). However, the actual number of such couples is 38.
Thus the population of this cell is disproportionately high. Note that the 38 mothers in this cell
constitute 46.3% of all 82 mothers who are Highly Supportive, while due to chance alone only
34% of Highly Supportive mothers should be in that cell, since that is the proportion of all
fathers who are Highly Supportive. Similarly, due to chance alone we would expect 12 of the 82
Mother and Father Parenting 22
Highly Supportive mothers to be paired with Unsupportive-Detached fathers, since 15% of all
fathers are Unsupportive-Detached (.15 * 82 = 12.3), but in fact only eight Highly Supportive
mothers are paired with Unsupportive-Detached fathers. Thus the population of this cell is
disproportionately low.
Overall, there was support for the hypothesis that parents would resemble each other with
respect to parenting pattern, χ2 (9, N = 200) = 20.48, p < .05. Highly Supportive parents were
disproportionately likely to be coupled together, as were Unsupportive-Negative parents.
However, Somewhat Supportive and Unsupportive-Detached parents were disproportionately
likely to be coupled with each other. The most striking finding was the symmetry of partner
pairings across sex. Unsupportive-Detached mothers were disproportionately likely to be paired
with Somewhat Supportive fathers, and Unsupportive-Detached fathers were disproportionately
likely to be paired with Somewhat Supportive mothers. Highly Supportive mothers were unlikely
to be paired with Somewhat Supportive, Unsupportive-Negative, or Unsupportive-Detached
fathers, while Highly Supportive fathers were unlikely to be paired with Unsupportive-Negative
or Unsupportive-Detached mothers.
Note that the disproportionality of a given cell’s count is based on the marginals of each
sex’s cluster distribution. Thus, although Somewhat Supportive mothers were disproportionately
likely to pair with Unsupportive-Detached fathers, they were most likely to pair with Somewhat
Supportive fathers, who constituted the largest group of fathers. Unsupportive (Negative and
Detached) mothers were also more likely to pair with Somewhat Supportive fathers than any
other cluster. However, Highly Supportive mothers were most likely to pair with Highly
Supportive fathers. Interestingly, although Highly Supportive mothers were the most populous
group, Unsupportive-Negative fathers were most likely to be paired with Unsupportive-Negative
Mother and Father Parenting 23
mothers, and Unsupportive-Detached fathers were most likely to be paired with Somewhat
Supportive mothers.
Associations Between Parenting Pairs and Child Cognitive Outcomes
To address our final two research questions, we used couples' cluster pairings to predict
child math and language scores at age five. Due to small cell sizes, it was necessary to
consolidate some parenting pattern combinations. Because there had been no significant
differences between the children of Highly and Somewhat Supportive mothers, those two
clusters were combined into a single Supportive group. The same was true of fathers, so the same
consolidation was performed for father clusters. Because there had been no significant
differences between children in the Unsupportive-Detached and Unsupportive-Negative mother
clusters, those two clusters were combined into a single Unsupportive group. This consolidation
was repeated for father clusters. The resulting typology classified children according to whether
their mother and father were both Supportive (n = 121), only their mother was (n = 31), only
their father was (n = 31), or neither was (n = 17). Children in these four groups were then
compared according to their math and language scores. Marginal means were estimated for the
groups after regressing each outcome on group membership and the covariates. All means are
calculated with covariates set to their mean values (Table 5).
Children with a Supportive mother and a Supportive father scored highest, and children
with no Supportive parents scored lowest. The spread between them was 16 points (107% of a
standard deviation) on math and 8.9 points (59% of a standard deviation) on language.
Children whose mother and father were both Supportive also scored 7.0 points higher on
math (p < .05) and 5.7 points higher on language (p < .05) than children with only a Supportive
mother. Though not significant, they also scored 6.0 points higher on math and 5.2 points higher
Mother and Father Parenting 24
on language than children with only a Supportive father. Children without any Supportive
parents scored lower than those with either a Supportive mother or a Supportive father, although
these differences did not achieve statistical significance at the p < .05 level.
The groups were then further consolidated by combining children with either a
Supportive mother or a Supportive father to answer our third research question regarding the
implications of number of Supportive parents. This consolidation allowed for a three-group
comparison: children with two Supportive parents vs. children with one Supportive parent vs.
children with no Supportive parents. These groups’ marginal means on math and language were
estimated based on regression models adjusting for all covariates.
We found significant differences on both math and language between children with two
Supportive parents and those with one (Figure 1). While children with two Supportive parents
had an average score of 96.0 on math, those with one Supportive parent had an average score of
89.5, a difference of 6.5 points or 43% of a standard deviation (p < .05). Children with two
Supportive parents had an average score of 97.3 on language, compared to 91.8 among children
with one Supportive parent, yielding a difference of 5.5 points or 37% of a standard deviation (p
< .05). The comparison between children with one and no Supportive parents produced mixed
results. On math, children with one Supportive parent scored 9.5 points (63% of a standard
deviation) higher than children with no Supportive parents (89.5 versus 80.0, p < .05). However,
they scored only 3.4 points higher on language (91.8 versus 88.4), which was not statistically
significant.
The next issue to be addressed was whether the benefit conferred by a Supportive parent
was moderated by the other parent's Supportiveness. A rough answer to this question is found by
comparing the difference in points between children with one and no Supportive parents (the
Mother and Father Parenting 25
benefit conferred by a Supportive parent when the other parent is not Supportive), on the one
hand, with the difference in points between those with two and one Supportive parents (the
benefit conferred by a Supportive parent when the other parent is also Supportive), on the other.
A simple inspection of these differences suggests variation by developmental domain. The
difference in math points between children with no and one Supportive parents (9.5) was greater
than the difference between children with one and two Supportive parents (6.5). It therefore
appeared that a Supportive parent conferred a greater increase in math points when the other
parent was Unsupportive. However, on language, the difference between children with no and
one Supportive parents (3.4) was smaller than the difference between children with one and two
(5.5), which suggested that a Supportive parent conferred a greater increase in language points
when the other parent was also Supportive.
Thus there appeared to be an interactive effect between two parents’ Supportiveness,
albeit one running in opposite directions for math and language. To formally test this possibility,
regression equations with an interaction term were run. First, math scores were regressed on an
indicator variable denoting the presence of a Supportive mother, an indicator variable denoting
the presence of a Supportive father, and an interaction term multiplying the two. All covariates
were included as controls. A second model was run for language scores.
The interaction term (Supportive mother * Supportive father) was insignificant in both
models (not shown). Therefore, although it appeared at first that a Supportive parent was
associated with a greater increase in math score when the other parent was Unsupportive, but a
greater increase in language score when the other parent was also Supportive, neither finding
was supported by tests of significance. Consequently, we cannot reject the null hypothesis that
two Supportive parents have an additive or linear effect. Thus for both math and language, the
Mother and Father Parenting 26
combination of a Supportive mother and a Supportive father results in an increase equivalent to
the sum of the increases associated with each Supportive parent separately. Otherwise stated, we
cannot conclude that the increase in points associated with a Supportive parent depends on
whether the other parent is also Supportive.
Associations Between the Sex of a Sole Supportive Parent and Child Cognitive Outcomes
By comparing children with only a Supportive mother to those with only a Supportive
father, we were able to answer our fourth research question regarding the significance of the sex
of a single Supportive parent. As shown in Table 5, among children with only one Supportive
parent, there were no significant differences in math and language according to the sex of that
parent. Children whose Supportive parent was their mother scored 1.0 point lower on math and
0.5 points lower on language than those whose Supportive parent was their father.
Discussion
Previous attention to parenting in two-parent families with young children has centered
on sex differences in parenting behaviors and their influence on child outcomes. For example,
fathers have been found to be less sensitive and stimulating (Belsky et al., 1984), engage in more
physical (MacDonald & Parke, 1984) and less instructive (Stevenson et al., 1988) play, and
prohibit their children’s actions more (Brachfeld-Child, 1986) than mothers. Entirely absent from
the literature is a consideration of the actual parent pairings in which children are likely to find
themselves, and the influence of parents’ combination of parenting styles on children’s
development.
Studies describing maternal and paternal parenting based on sex-specific samples are of
limited utility. The rigors of statistical analysis demand that a determination of the effect of one
parent’s parenting pattern on a child account for the other parent’s pattern, due to the possibility
Mother and Father Parenting 27
that the putative effects of one parent’s pattern are in fact driven by the other’s or by a
synergistic dynamic created by a combination of the two. Studies that include mothers and
fathers in a single model are an improvement over single-sex studies, but if the model attempts to
isolate the effect of each parent’s parenting behavior by controlling for the other’s it is likely to
be inadequate for two reasons. First, there may be nonrandom configurations of parent couplings.
If certain parent pairings are more prevalent than others, it should be fruitful to consider a
parent’s parenting pattern in the context of the other parent’s pattern with which it is likely to be
found. Second, there may be interactive effects of couples’ parenting on children that are not
detectable in studies that examine each parent individually.
Additionally, the consideration of only one parent at a time can lead to inaccurate
predictions for children in two-parent families. As an illustration, consider a hypothetical
analysis limited to associations between maternal supportiveness and child cognition which finds
that a child with a supportive mother scores five points higher on average than a child without
one. However, among children with a supportive mother, those whose father is also supportive
may score eight points higher than a child without a supportive mother, while those whose father
is not supportive may score only three points higher. In such a scenario, the failure to account for
the father’s supportiveness would generate less than accurate predictions about both groups –
those with and those without a supportive father.
It is incumbent upon parenting scholars to devise ways of depicting children's parenting
experiences in two-parent families that reflect the full complexity of the family. This will likely
require experimentation with different measures of joint parental influence. The present study
classifies mothers and fathers separately into groups according to their behavior in separate
dyadic interactions with their child, and then examines couples according to the cross-
Mother and Father Parenting 28
classification of their groups. Future efforts to describe couples should observe both parents
simultaneously in order to generate more nuanced classifications of their parenting behavior. In
addition to varying the data collection procedure, future studies should try alternative variable
types (e.g., continuous scales vs. categorical groups) and statistical procedures (e.g., structural
equation modeling vs. regression). Finally, larger (and possibly more diverse) samples of parents
than the one examined here would allow for more powerful comparisons among combinations of
parenting styles within couples and their effects on child development. It will be necessary to
draw on multiple samples and methodologies to substantively advance the body of knowledge
about parenting processes in two-parent families.
To summarize the present study's results, our first research question asked whether
maternal and paternal parenting at age two each predicted child cognitive outcomes at age five
adjusting for age two cognitive outcomes. We find that both mothers' and fathers' parenting at
age two predicts children's math and language at five. The most salient distinction for both
mothers and fathers was between Supportiveness vs. Unsupportiveness. With respect to our
second research question, our hypothesis that the coresident couples in our sample would exhibit
similar parenting patterns found only mixed support. Both Highly Supportive and Negative
parents were disproportionately likely to be paired with others in their cluster. However,
Somewhat Supportive and Detached parents were also disproportionately likely to be paired. It is
difficult to draw conclusions about this finding because the direction of the association between
mating and parenting pattern is unknown. It may be that people who tend to be Somewhat
Supportive and Detached parents are drawn to each other; it may also be the case that in couples
in which both members have a tendency to be Detached, one is likely to evolve into a Somewhat
Supportive parent for the benefit of the child. Conversely, in couples in which both members
Mother and Father Parenting 29
have a tendency to be Somewhat Supportive, one may be likely to fall into a pattern of
Detachment.
In answer to our third question, we fail to find significant interactive effects between
maternal and paternal supportiveness in a sample of low-income families with young children.
Therefore, we cannot conclude that the advantage for a young child’s cognitive development
associated with having a supportive parent turns on the other parent’s parenting behaviors. Our
previous analysis (Ryan et al., 2006) also failed to identify a significant interaction between
mothers' and fathers' supportiveness on child cognitive ability at ages two and three. At age two,
it appeared that a Supportive parent was particularly beneficial when the other parent was
Unsupportive; however, by age three the benefit of a Supportive parent was the same regardless
of whether the other parent was Unsupportive or Supportive. Although the interpretability of the
pattern of findings across ages two, three, and five is limited by the discrepancy in measures
between ages two and three, on the one hand, and age five, on the other, it nonetheless seems
prudent to conclude that this sample suggests an additive effect of parental supportiveness on
young children's cognitive status, contrary to our hypothesis. Future research should attempt to
replicate these results with other samples.
In response to our fourth and last research question, we find no evidence that the sex of a
supportive parent influences math or language scores at age five. This result is consistent with
those from child ages two and three (Ryan et al., 2006). However, caution must be taken in
interpreting these findings in light of the selectiveness of the present sample. We estimate that
the couples in this sample represent less than half of all cohabiting couples in the larger study
(based on maternal report of cohabitation). Fathers who agreed to participate in the substudy of
fathers may be more involved fathers in general. Their willingness to participate may signal a
Mother and Father Parenting 30
greater commitment to or confidence in their role as a father, either of which is likely to
influence their parenting style. Therefore, the generalizability of our finding to all low-income
families may be limited.
Furthermore, the degree to which our findings with a low-income sample are
generalizable to families with higher incomes is unknown, given that our study is the first to
formally examine interactive effects between mothers and fathers, and to cross-classify parents
within couples, in any population. However, the literatures independently establishing
associations between maternal and paternal early supportiveness and favorable cognitive
outcomes in young children have drawn on both low- and middle-class samples (Black et al.,
1999; Cowan et al., 1991; Kelley et al., 1998; Landry et al., 2001; Zaslow et al., 2006). Further
research is needed to investigate whether the joint influence of mothers’ and fathers’ parenting
behaviors on young children’s development varies according to socioeconomic status.
Despite the selective nature of the sample, these findings are relevant to a discussion of
low-income families with two involved parents. Such families are often the targets, and most
engaged consumers, of intervention services like Early Head Start. Our findings regarding the
advantage of having a supportive parent regardless of parent gender suggest that early
interventions to improve parenting should target resident fathers as vigorously as mothers, and
that practitioners should address both parents’ behaviors in concert. To further inform early
intervention programs as well as early childhood research about the nature and consequences of
children’s early parenting environments, it is hoped that future studies of child development in
two-parent families will continue to explore the joint contributions made by mothers and fathers.
Mother and Father Parenting 31
References
Ainsworth, M. D., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: A
psychological study of the strange situation. Hillsdale, NJ: Erlbaum.
Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster analysis. Newbury Park: Sage
Publications.
Baumrind, D. (1967). Child care practices anteceding three patterns of preschool behavior.
Genetic Psychology Monographs, 1, 43-88.
Bayley, N. (1993). Bayley Scales of Infant Development: Manual (2nd ed.). New York:
Psychological Corporation, Harcourt, Brace & Company.
Belsky, J., & Fearon, R. M. P. (2004). Exploring marriage-parenting typologies and their
contextual antecedents and developmental sequelae. Development and Psychopathology,
16, 501-523.
Belsky, J., Gilstrap, B., & Rovine, M. (1984). The Pennsylvania Infant and Family Development
Project, I: Stability and change in mother-infant and father-infant interaction in a family
setting at one, three, and nine months. Child Development, 55, 692-705.
Bentley, K. S., & Fox, R. A. (1991). Mothers and fathers of young children: Comparison of
parenting styles. Psychological Reports, 69, 320-322.
Berk, L. E., & Winsler, A. (1995). Scaffolding children's learning: Vygotsky and early childhood
education. Washington, DC: National Association for the Education of Young Children.
Black, M. M., Dubowitz, H., & Starr, R. H. (1999). African American fathers in low income,
urban families: Development, behavior, and home environment of their three-year-old
children. Child Development, 70, 967-978.
Mother and Father Parenting 32
Blackwell, D. L., & Lichter, D. L. (2000). Mate selection among married and cohabiting couples.
Journal of Family Issues, 21, 275-302.
Bowlby, J. (1982). Attachment and loss. Vol 1. Attachment (2nd ed.). New York: Basic Books.
Brachfield-Child, S. (1986). Parents as teachers: Comparisons of mothers’ and fathers’
instructional interaction with infants. Infant Behavior & Development, 9, 127-131.
Brady-Smith, C., O’Brien, C., Berlin, L., & Ware, A. (1999). 24-Month child-parent interaction
Rating scales for the Three-Bag assessment. New York: Teachers College, Columbia
University.
Brody, G. H., Pillegrini, A. D., & Sigel, I. E. (1986). Marital quality and mother-child and father-
child interactions with school-aged children. Developmental Psychology, 22, 291-296.
Conner, D. B., Knight, D. K., & Cross, D. R. (1997). Mothers’ and fathers’ scaffolding of their
2-year-olds during problem-solving and literacy interactions. British Journal of
Developmental Psychology, 15, 323-338.
Cowan, C. P., Cowan, P. A., Heming, G., & Miller, N. B. (1991). Becoming a family: Marriage,
parenting, and child development. In P. A. Cowan & M. Hetherington (Eds.), Family
transitions (pp. 79-109). Hillsdale, N. J.: Erlbaum.
Darling, N., & Steinberg, L. (1993). Parenting style as context: An integrative model.
Psychological Bulletin, 113, 487-496.
Dierker, L. C., Merikangas, K., R., & Szatmari, P. (1999). Influence of parental concordance for
psychiatric disorders on psychopathology in offspring. Journal of the American Academy
of Child and Adolescent Psychiatry, 38, 280-288.
Duncan, G. J., Claessens, A., & Engel, M. (2005, April). School readiness and third grade
achievement: Which cognitive skills and socioemotional behaviors matter most? Paper
Mother and Father Parenting 33
presented at the biennial meeting of the Society for Research in Child Development,
Atlanta, GA.
Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocabulary Test – Third Edition. Circle
Pines, MN: American Guidance Service, Inc.
Elder, G. H. (1998). The life course as developmental theory. Child Development, 69, 1-12.
Ensminger, M. E., & Slusarcick, A. L. (1992). Paths to high school graduation or dropout: A
longitudinal study of a first-grade cohort. Sociology of Education, 65, 95-113.
Erel, O., & Burman, B. (1995). Interrelatedness of marital relations and parent-child relations: A
meta-analytic review. Psychological Bulletin, 118, 108-132.
Galbaud du Fort, G., Bland, R. C., Newman, S. C., & Boothroyd, L. J. (1998). Spouse similarity
for lifetime psychiatric history in the general population. Psychological Medicine, 28,
789-803.
Goodman, S. H., Brogan, D., Lynch, M. E., & Fielding, B. (1993). Social and emotional
competence in children of depressed mothers. Child Development, 64, 516-531.
Goodman, S. H., & Gotlib, I. H. (1999). Risk for psychopathology in the children of depressed
mothers: A developmental model for understanding mechanisms of transmission.
Psychology Review, 106, 458-490.
Gauvain, M., Fagot, B. I., Leve, C., & Kavanagh, K. (2002). Instruction by mothers and fathers
during problem solving with their young children. Journal of Family Psychology, 16, 81-
90.
Hart, C. H., DeWolf, D. M., Wozniak, P., & Burts, D. C. (1992). Maternal and paternal
disciplinary styles: Relations with preschoolers’ playground behavioral orientations and
peer status. Child Development, 63, 879-892.
Mother and Father Parenting 34
Henry, D. B., Tolan, P. H., & Gorman-Smith, D. (2005). Cluster analysis in family psychology
research. Journal of Family Psychology, 19, 121-132.
Kahn, R. S., Brandt, D., & Whitaker, R.C. (2004). Combined effect of mothers’ and fathers’
mental health symptoms on children’s behavioral and emotional well-being. Archives of
Pediatric and Adolescent Medicine, 158, 721-729.
Kalmijn, M. (1998). Intermarriage and homogamy: Causes, patterns, trends. Annual Review of
Sociology, 24, 395-421.
Kelley, M. L., Smith, T. S., Green, A. P., Berndt, A. E., & Rogers, M. C. (1998). Importance of
fathers’ parenting to African-American toddler’s social and cognitive development.
Infant Behavior & Development, 21, 733-744.
Kerig, P. K., Cowan, P. A., & Cowan, C. P. (1993). Marital quality and gender differences in
parent-child interaction. Developmental Psychology, 29, 931-939.
Krishnakumar, A., & Buehler, C. (2000). Interparental conflict and parenting behaviors: A meta-
analytic review. Family Relations, 49, 25-43.
Krueger, R. F., Moffitt, T. E., Caspi, A., Bleske, A., & Silva, P. A. (1998). Assortative mating
for antisocial behavior: Developmental and methodological implications. Behavior
Genetics, 28, 173-186
Lamb, M. (1997). Fathers and child development: An introductory overview. In M. Lamb (Ed.),
The role of the father in child development (3rd ed.) (pp. 1-18). New York: John Wiley &
Sons, Inc.
Landry, S. H., Smith, K.E., Swank, P. R., Assel, M. A., & Vellet, S. (2001). Does early
responsive parenting have a special importance for children’s development or is
consistency across early childhood necessary? Developmental Psychology, 37, 387-403.
Mother and Father Parenting 35
Lovas, G. S. (2005). Gender and patterns of emotional availability in mother-toddler and father-
toddler dyads. Infant Mental Health Journal, 26, 327-353.
Love, J. M., Kisker, E. E., Ross, C. M., Schochet, P. Z., Brooks-Gunn, J., Paulsell, D., et al.
(2002). Making a difference in the lives of infants and toddlers in their families: The
impacts of Early Head Start (Vol. 2). Washington, DC: Administration on Children,
Youth and Families, U. S. Department of Health and Human Services.
Lovejoy, M. C., Graczyk, P. A., O’Hare, E., & Neuman, G. (2000). Maternal depression and
parenting behavior: A meta-analytic review. Clinical Psychology Review, 20, 561-592.
Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the family: Parent-child
interaction. In P. H. Mussen (Ed.) & E. M. Hetherington (Vol. Ed.), Handbook of child
psychology: Vol. 4. Socialization, personality, and social development (4th ed.)(pp. 1-
101). New York: Wiley.
MacDonald, K., & Parke, R. D. (1984). Bridging the gap: Parent-child play interaction and peer
interactive competence. Child Development, 55, 1265-1277.
McLeod, J. D., & Nonnemaker, J. M. (2000). Poverty and child emotional and behavioral
problems: Racial/ethnic differences in processes and effects. Journal of Health and
Social Behavior, 41, 137-161.
McLoyd, V. C. (1990). The impact of economic hardship on black families and children:
Psychological distress, parenting, and socioemotional development. Child Development,
61, 311-346.
McLoyd, V. C. (1998). Socioeconomic disadvantage and child development. American
Psychologist, 53, 185-204.
Mother and Father Parenting 36
McLoyd, V. C., Jayaratne, T. E., Ceballo, R., & Borquez, J. (1994). Unemployment and work
interruption among African American single mothers: Effects on parenting and
adolescent socioemotional functioning. Child Development, 65, 562-589.
Minuchin, P. (1985). Families and individual development: Provocations from the field of family
therapy. Child Development, 56, 289-302.
NICHD Early Child Care Research Network. (1999). Chronicity of maternal depressive
symptoms, maternal sensitivity, and child functioning at 36 months. Developmental
Psychology, 35, 1297-1310.
NICHD Early Child Care Research Network. (2004). Fathers’ and mothers’ parenting behavior
and beliefs as predictors of children’s social adjustment in the transition to school.
Journal of Family Psychology, 18, 628-638.
Pettit, G. S., Bates, J. E., & Dodge, K. A. (1997). Supportive parenting, ecological context, and
children’s adjustment: A seven-year longitudinal study. Child Development, 68, 908-923.
Pleck, J. H. (1997). Paternal involvement: Levels, sources, and consequences. In M. Lamb (Ed.),
The role of the father in child development (3rd ed.) (pp.66-103). New York: John Wiley
& Sons, Inc.
Power, T. G., & Parke, R. D. (1983). Patterns of mother and father play with their 8-month-old
infant: A multiple analyses approach. Infant Behavior & Development, 8, 453-459.
Pratt, M.W., Kerig, P., Cowan, P.A., & Cowan, C.P. (1988). Mothers and fathers teaching 3-
year-olds: Authoritative parenting and adult scaffolding of young children’s learning.
Developmental Psychology, 24, 832–839.
Qian, Z. (1997). Breaking the racial barriers: Variations in interracial marriage between 1980 and
1990. Demography, 34, 263-276.
Mother and Father Parenting 37
Rogoff, B., Mistry, J., Goncu, A., & Mosier, C. (1993). Guided participation in cultural activity
by toddlers and caregivers. Monographs of the Society for Research in Child
Development, 58 (8, Serial No. 236).
Ryan, R. M., Martin, A., & Brooks-Gunn, J. (2006). Is one good parent good enough? Patterns of
mother and father parenting and child cognitive outcomes at 24 and 36 months.
Parenting: Science and Practice, 6, 211-228.
Schoen, R., & Weinick, R. M. (1993). Partner choice in marriages and cohabitations. Journal of
Marriage and the Family, 55, 408-414.
Shaywitz, B., Holford, T., Holahan, J., Fletcher, J., Stuebing, K., Francis, D., et al. (1995). A
Matthew effect for IQ but not for reading: Results from a longitudinal study. Reading
Research Quarterly, 30, 894-906.
Steinberg, L., Lamborn, S. D., Darling, N., Mounts, N. S., & Dornbusch, S. M. (1994). Over-
time changes in adjustment and competence among adolescents from authoritative,
authoritarian, indulgent, and neglectful families. Child Development, 65, 754-770.
Stevenson, M. B., Leavitt, L. A., Thompson, R. H., & Roach, M. A. (1988). A social relations
model analysis of parent and child play. Developmental Psychology, 24, 101-108.
Stevenson, H. W., & Newman, R. S. (1986). Long-term prediction of achievement and attitudes
in mathematics and reading. Child Development, 57, 646-659.
Tamis-LeMonda, C. S., Bornstein, M. H., & Baumwell, L. (2001). Maternal responsiveness and
children’s achievement of language milestones. Child Development, 72, 748-767.
Tamis-LeMonda, C. S., Shannon, J. D., Cabrera, N. J., & Lamb, M. E. (2004). Fathers and
mothers at play with their 2- and 3-year-olds: Contributions to language and cognitive
development. Child Development, 75, 1806-1820.
Mother and Father Parenting 38
Van Bakel, H. J. A., & Riksen-Walraven, J. M. (2002). Parenting and development of one-year-
olds: Links with parental, contextual, and child characteristics. Child Development, 73,
256-273.
Vandell, D. L. (1979). A microanalysis of toddlers' social interaction with mothers and fathers.
Journal of Genetic Psychology, 134, 299-312.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes.
Cambridge, MA: Harvard Press.
Williamson, G., Appelbaum, M., & Epanchin, A. (1991). Longitudinal analyses of academic
achievement. Journal of Educational Measurement, 28, 61-76.
Winsler, A., Madigan, A. L., & Aquilino, S. A. (2005). Correspondence between maternal and
paternal parenting styles in early childhood. Early Childhood Research Quarterly, 20, 1-
12.
Woodcock, R., & Johnson, M. (1989). Woodcock-Johnson Psycho-Educational Test Battery –
Revised. Chicago, IL: Riverside Publishing Company.
Zaslow, M. J., Weinfeld, N. S., Gallagher, M., Hair, E. C., Ogawa, J. R., Egeland, B., et al.
(2006). Longitudinal prediction of child outcomes from differing measures of parenting
in a low-income sample. Developmental Psychology, 42, 27-37.
Mother and Father Parenting 39
Author Note
Rebecca M. Ryan is now at the Harris School of Public Policy Studies, University of
Chicago.
The findings reported here are based on research conducted as part of the national Early
Head Start Research and Evaluation Project funded by the Administration for Children and
Families (ACF), U.S. Department of Health and Human Services under contract 105-95-1936 to
Mathematica Policy Research, Princeton, NJ, and Columbia University’s National Center for
Children and Families, Teachers College in conjunction with the Early Head Start Research
Consortium, and under Task Order No. 32, Contract 282-98-0021, to Mathematica Policy
Research and 15 university research partners. The Consortium consists of representatives from
17 programs participating in the evaluation, 15 local research teams, the evaluation contractors,
and ACF. Research institutions in the Consortium (and principal researchers for conducting this
research through 36 months of age) include ACF (Rachel Chazan Cohen, Judith Jerald, Esther
Kresh, Helen Raikes, and Louisa Tarullo); Catholic University of America (Michaela Farber,
Harriet Liebow, Nancy Taylor, Elizabeth Timberlake, and Shavaun Wall); Columbia University
(Lisa Berlin, Christy Brady-Smith, Jeanne Brooks-Gunn, and Allison Sidle Fuligni); Harvard
University (Catherine Ayoub, Barbara Alexander Pan, and Catherine Snow); Iowa State
University (Dee Draper, Gayle Luze, Susan McBride, Carla Peterson); Mathematica Policy
Research (Kimberly Boller, Jill Constantine, Ellen Eliason Kisker, John M. Love, Diane Paulsell,
Christine Ross, Peter Schochet, Cheri Vogel, and Welmoet van Kammen); Medical University of
South Carolina (Richard Faldowski, Gui-Young Hong, and Susan Pickrel); Michigan State
University (Hiram Fitzgerald, Tom Reischl, and Rachel Schiffman); New York University (Mark
Spellmann and Catherine Tamis-LeMonda); University of Arkansas (Robert Bradley, Richard
Mother and Father Parenting 40
Clubb, Andrea Hart, Mark Swanson, and Leanne Whiteside-Mansell); University of California,
Los Angeles (Carollee Howes and Claire Hamilton); University of Colorado Health Sciences
Center (Robert Emde, Jon Korfmacher, JoAnn Robinson, Paul Spicer, and Norman Watt);
University of Kansas (Jane Atwater, Judith Carta; and Jean Ann Summers); University of
Missouri-Columbia (Mark Fine, Jean Ispa, and Kathy Thornburg); University of Pittsburgh (Beth
Green, Carol McAllister, and Robert McCall); University of Washington School of Education
(Eduardo Armijo and Joseph Stowitschek); University of Washington School of Nursing
(Kathryn Barnard and Susan Spieker), and Utah State University (Lisa Boyce, Gina Cook,
Catherine Callow-Heusser, and Lori Roggman). The content of this publication does not
necessarily reflect the views or policies of the Department of Health and Human Services.
Portions of this article were presented at the biennial meeting of the Society for Research
in Child Development, Atlanta, Georgia, April 2005.
Mother and Father Parenting 41
Table 1.
Intercorrelations Among Maternal and Paternal Parenting Scales
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
1. Maternal
sensitivity
-- .72* .73* -.43* -.50* -.45* .32* .21* .26* -.20* -.11 -.30*
2. Maternal
cognitive
stimulation
-- .55* -.18* -.52* -.23*
.25* .19* .23* -.04 -.13 -.20*
3. Maternal
positive regard
-- -.27* -.41* -.28* .27* .10 .26* -.14 -.03 -.18*
4. Maternal
intrusiveness
-- -.05 .38* -.17* -.16* -.12 .25* .05 .12
5. Maternal
detachment
-- .15* -.17* -.09 -.12 .02 .16* .12
6. Maternal
negative regard
-- -.23* -.18* -.14 .25* -.01 .35*
Mother and Father Parenting 42
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
7. Paternal
sensitivity
-- .57* .67* -.48* -.47* -.49*
8. Paternal
cognitive
stimulation
-- .54* -.10 -.41* -.17*
9. Paternal
positive regard
-- -.24* -.37* -.31*
10. Paternal
intrusiveness
-- .04 .51*
11. Paternal
detachment
-- .25*
12. Paternal
negative regard
--
Mother and Father Parenting 43
Table 2
Parenting Scale Scores by Mother and Father Parenting Cluster
Mothers Fathers
Highly Somewhat Unsupportive - Unsupportive -
Highly Somewhat Unsupportive - Unsupportive -
Supportive Supportive Negative Detached F
Supportive Supportive Negative Detached F
(n = 82) (n = 70) (n = 29) (n = 19) (n = 68) (n = 84) (n = 18) (n = 30)
Sensitivity 5.7 (0.6)a 4.6 (0.5)b 3.7 (0.8)c 3.3 (0.7)c 126.0*** 5.6 (0.6)a 4.6 (0.6)b 3.2 (0.6)c 3.5 (0.7)c 124.3***
Cognitive
stimulation 5.0 (0.8)a 3.8 (0.8)b 3.7 (0.8)b 2.2 (0.6)c 81.5*** 4.8 (0.8)a 3.9 (0.7)b 3.5 (0.8)b 2.8 (0.5)c 54.2***
Positive
regard 5.1 (0.8)a 3.4 (0.7)b 2.9 (1.0)c 2.2 (0.7)d 116.7*** 5.0 (0.8)a 3.3 (0.8)b 2.4 (0.7)c 1.9 (0.5)c 157.8***
Negative
regard 1.1 (0.3)a 1.2 (0.5)ab 2.1 (0.9)c 1.6 (1.2)bc 18.6*** 1.0 (0.2)a 1.2 (0.4)ab 2.4 (0.7)c 1.4 (0.7)b 43.9***
Intrusiveness 1.4 (0.6)a 1.5 (0.6)a 3.5 (0.8)b 1.4 (0.7)a 93.3*** 1.5 (0.6)a 1.8 (0.7)a 3.9 (0.8)b 1.7 (0.8)a 54.4***
Detachment 1.1 (0.2)a 1.2 (0.5)a 1.3 (0.5)a 2.8 (1.0)b 74.4*** 1.0 (0.1)a 1.2 (0.4)ab 1.4 (0.8)b 1.9 (0.9)c 23.8***
Note. Table presents M (SD) with test statistic computed via ANOVA; N = 200; Within-sex means with different subscripts differ
significantly at p < .05 in post-hoc comparisons with Bonferroni adjustments.
*** p < .001.
Mother and Father Parenting 44
Table 3
Child Math and Language Scores at Age Five by Mother and Father Parenting Cluster
Mothers
Fathers
Highly Somewhat Unsupportive- Unsupportive-
Highly Somewhat Unsupportive- Unsupportive-
Supportive Supportive Negative Detached Supportive Supportive Negative Detached
Math 95.6 a 93.2
ab 87.1
b 85.8
b 94.9
a 94.6
a 84.2
b 87.3
b
Language 97.8
a 94.1
ab 91.7
ab 89.2
b 97.8
a 94.9
ab 90.4
ab 90.7
b
Note. Means adjusted for maternal race, program status, highest parental education, paternal biological status, poverty status, and age
two mental development (Bayley, 1993). Within-sex means with different subscripts differ significantly at the p < .05 level in
Bonferroni-adjusted comparisons. Math assessed with the Woodcock-Johnson-Revised Applied Problems subtest (Woodcock &
Johnson, 1990; n = 194). Language assessed with the Peabody Picture Vocabulary Test-III (Dunn & Dunn, 1997; n = 191).
Mother and Father Parenting 45
Table 4
Cross-Classification of Mother and Father Parenting Clusters
Fathers
Mothers
Highly
Supportive
Somewhat
Supportive
Unsupportive-
Negative
Unsupportive-
Detached
Total
Highly Supportive
Observed count
Expected count
Row percentage
Column percentage
38
28
46.3
55.9
31
34
37.8
36.9
5
8
6.1
27.8
8
12
9.8
26.7
82
82
100.0
41.0
Somewhat Supportive
Observed count
Expected count
Row percentage
Column percentage
23
24
32.9
33.8
29
29
41.4
34.5
4
6
5.7
22.2
14
11
20.0
46.7
70
70
100.0
35.0
Mother and Father Parenting 46
Unsupportive - Negative
Observed count
Expected count
Row percentage
Column percentage
5
10
17.2
7.4
13
12
44.8
15.5
6
3
20.7
33.3
5
4
17.2
16.7
29
29
100.0
14.5
Unsupportive - Detached
Observed count
Expected count
Row percentage
Column percentage
2
6
10.5
2.9
11
8
57.9
13.1
3
2
15.8
16.7
3
3
15.8
10.0
19
19
100.0
9.5
Total
Observed count
Expected count
Row percentage
Column percentage
68
68
34.0
100.0
84
84
42.0
100.0
18
18
9.0
100.0
30
30
15.0
100.0
200
200
100.0
100.0
Note. Chi-square (9 df, N = 200) = 20.48, p < .05.
Mother and Father Parenting 47
Table 5
Child Math and Language Scores at Age Five by Parental Supportiveness Pairing
Note. Means adjusted for maternal race, program status, highest parental education, paternal
biological status, poverty status, and age two mental development (Bayley, 1993). Within-row
means with different subscripts differ significantly at the p <. 05 level in Bonferroni-adjusted
comparisons. Math assessed with the Woodcock-Johnson-Revised Applied Problems subtest
(Woodcock & Johnson, 1990; n = 194). Language assessed with the Peabody Picture Vocabulary
Test-III (Dunn & Dunn, 1997; n = 191).
Mother
Supportive,
Father
Supportive
Mother
Supportive,
Father
Unsupportive
Mother
Unsupportive,
Father
Supportive
Mother
Unsupportive,
Father
Unsupportive
Math
96.0
a
89.0
b
90.0
ab 80.0 bc
Language 97.3
a
91.6 b 92.1
ab 88.4 b
Mother and Father Parenting 48
Figure Caption
Figure 1. Child Math and Language Scores at Age Five by Number of Supportive Parents.
Means adjusted for maternal race, program status, highest parental education, paternal biological
status, poverty status, and age two mental development (Bayley, 1993). Means with different
subscripts differ significantly at the p < .05 level in Bonferroni-adjusted comparisons. Math
assessed with the Woodcock-Johnson-Revised Applied Problems subtest (Woodcock & Johnson,
1990; n = 194). Language assessed with the Peabody Picture Vocabulary Test III (Dunn & Dunn,
1997; n = 191).
Mother and Father Parenting - Figure
0
20
40
60
80
100
120
Math Language
Standardized Score
0 Supportive parents
1 Supportive parent
2 Supportive parents
80.0a 89.5b 96.0c 88.4a 91.8a 97.3b
