Correlates of Parental Differential Treatment

Record: 1
Correlates of Parental Differential Treatment: Parental and
Contextual Factors During Middle Childhood.
Atzaba‐Poria, Naama (AUTHOR)
Pike, Alison (AUTHOR)
Child Development. Jan/Feb2008, Vol. 79 Issue 1, p217-232. 16p.
5 Charts, 4 Graphs.
Article
*AFFECT (Psychology)
*ADJUSTMENT (Psychology) in children
*PARENTAL acceptance
*CORRELATION (Statistics)
*CONTEXT effects (Psychology) in children
*FAMILIES
The current study examined whether parental and contextual risk
factors contribute to mothers’ and fathers’ differential treatment
(MDT/FDT) when accounting for sibling dyad characteristics. Also
explored was whether family type (single mothers vs. 2 parents)
moderated the links between the parental and contextual correlates
and MDT. One hundred and seventy-two families with older ( M=
7.4 years) and younger ( M= 5.2 years) siblings were studied.
Parents and children reported about the parent–child relationship,
and parents reported about the children’s characteristics, their own
psychological resources, and contextual factors. Controlling for
sibling dyad characteristics, FDT was predicted most consistently
by household chaos. Furthermore, single mothers were not at risk
per se for using more MDT but only when coupled with high
maternal anger. [ABSTRACT FROM AUTHOR]
Copyright of Child Development is the property of Wiley-Blackwell
and its content may not be copied or emailed to multiple sites or
posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles
for individual use. This abstract may be abridged. No warranty is
given about the accuracy of the copy. Users should refer to the
original published version of the material for the full abstract.
(Copyright applies to all Abstracts.)
Ben‐Gurion University
University of Sussex
10352
1
2
1
2
ISSN:
DOI:
Accession Number:
Database:
0009-3920
10.1111/j.1467-8624.2007.01121.x
28794108
Academic Search Complete
Correlates of Parental Differential Treatment: Parental and Contextual Factors During
Middle Childhood.
The current study examined whether parental and contextual risk factors contribute to mothers' and
fathers' differential treatment (MDT/FDT) when accounting for sibling dyad characteristics. Also explored
was whether family type (single mothers vs. 2 parents) moderated the links between the parental and
contextual correlates and MDT. One hundred and seventy‐two families with older (M= 7.4 years) and
younger (M= 5.2 years) siblings were studied. Parents and children reported about the parent–child
relationship, and parents reported about the children's characteristics, their own psychological resources,
and contextual factors. Controlling for sibling dyad characteristics, FDT was predicted most consistently
by household chaos. Furthermore, single mothers were not at risk per se for using more MDT but only
when coupled with high maternal anger.
Twenty years ago, [55] alerted developmental psychologists to the fact that children within the same
family often turn out quite differently and that sharing the same rearing environment does not lead to
sibling similarity. This began a wave of research to pinpoint child‐specific aspects of the environment
linked to sibling differences in outcome, and parental differential treatment (PDT) has emerged as a
consistent factor related to children's adjustment ([63]). There have been very few studies, however,
examining risk factors for PDT. The aim of the current study was to assess how parental characteristics
and contextual factors are associated with PDT, while controlling for differing characteristics of the sibling
dyad. We examined this issue during middle childhood using a multi‐informant approach.
Theoretical Frameworks
The most influential theoretical framework for considering correlates of parenting was formulated by [ 4].
Reviewing research from abusive families, Belsky postulated that child characteristics, personal
resources of the parents, and contextual sources are all important determinants of parenting. In the
current study, it is not parenting, per se, that is of interest but rather the degree of differential treatment
that parents exhibit. We hypothesize, however, that parental and contextual factors may also be
associated with differential treatment, considering that the division of attention and affection between
children is at risk under more stressful circumstances and/or for parents with less psychological
resources:
The underlying notion is that parents have a finite amount of resources in terms of time, attention,
patience and support to give their children. In families in which most of these resources are devoted to
coping with economic stress, depression, and/or marital conflict, parents may become less consciously or
intentionally equitable and more driven by preferences or child characteristics in their childrearing efforts.
(Henderson, Hetherington, Mekos, & Reiss, 1996 , p. 47)
Treating children differently may reflect appropriate, sensitive parenting ([37]), as happens when parents
adjust their behavior to each child's needs and characteristics. Therefore, PDT is expected and seen to
be normative when it is equitable and related to children's own characteristics such as age and
temperamental differences ([10]) or special needs ([45]). However, the extant evidence on PDT is
consistent in showing that PDT is a negative phenomenon for the disfavored child. For example, PDT has
been related to children's self‐esteem ([46]), adjustment ([27], [29], [44]), and negative emotionality ([10]).
Therefore, it is important for risk factors to be identified. In the current study, we minimized "justified" PDT
by stipulating that the siblings be close in age, during middle childhood, and without any known
disabilities or special educational needs.
PDT and Children's Characteristics
Research has shown that both parents and children perceive PDT ([ 8], [18]). Furthermore, there is
evidence to suggest that children's perceptions and understanding of differential treatment are more
important than objective levels of PDT ([37]). The current study utilized a puppet interview technique
([47]), enabling the children's perceptions to be taken into account. This technique provides a tool for
gaining consistent and meaningful reports from children as young as 4 years of age. We were therefore
able to examine correlates of PDT as reported by the children as well as by the parents themselves.
Children's own characteristics, such as gender, age, and temperament, and especially sibling differences
in these characteristics, have been related to PDT. Mixed results have been found for the gender
composition of the pair and PDT. For example, mothers and fathers reported having more conflictual
relationships with their children from same‐gender dyads than with children from opposite‐gender dyads
([61]). It has been suggested that parents may expect children who are the same gender to behave
similarly, and thus, they might develop more conflictual relationships with the less well‐behaved child. In
addition, they may have a greater need to differentiate between children who are the same gender than
between those who are different genders. On the other hand, [17] suggested that parents have an easier
time recognizing and admitting to differential treatment when their two children are obviously different
(e.g., a boy and a girl). Overall, the findings regarding sex constellation are underwhelming, with many
studies reporting no differences (e.g., [11], [43]).
Sibling age gap has been more consistently related to PDT ([26]). For example, it has been found that
younger children enjoy more warmth and involvement with their parents ([ 9]). Most strikingly, [35]
reported that child age was the strongest child‐specific predictor of parental differential positivity and
negativity. In addition, although parents vary in the degree to which their behavior is affected by child
temperament ([35]), children temperament, especially negative affect, is known to evoke elevated levels
of negative parental behavior and to decrease positive parental behaviors ([ 3], [35]). Furthermore, [10]
found that differences in siblings' negative emotionality levels were more strongly linked with PDT than
was each child's absolute level of negative emotionality.
While controlling for sibling dyad variables (i.e., sibling age gap, sibling sex constellation, and differential
sibling temperament), we focused on shared family factors as correlates of PDT. We did this in order to
avoid simply capturing variance in PDT due to the siblings' distinct characteristics.
Parent and Context Factors as Correlates of PDT
Extensive research has examined the associations between parental characteristics (e.g., [56]), as well
as contextual variables and parenting (e.g., [66]), including research stemming from our project detailing
associations during the middle childhood period ([52]). Much less research has examined how such
factors relate to PDT—we propose that parental and contextual variables may also prove to be important
correlates of PDT. Thus, the current study utilizes absolute differences in parenting to assess PDT
because we were concerned with differentiating those families displaying small versus large amounts of
PDT rather than examining the within‐family effects of PDT.
Previous studies examining links between parental stress and PDT have focused on variables such as
neuroticism ([25]) and aspects of maternal mental health such as depression ([32], [62]), as well as
contextual stress variables such as marital conflict (e.g., [19], [32], [35], [44]) and economic pressure
([32], [35], [48]). However, most of these studies have focused on only one or two stressors at a time (cf.
[32], [35]), ignoring more complex patterns of influence. We sought to take a more comprehensive
approach to reveal how different variables together are related to PDT. The one recent study to examine
multiple stressors and PDT found that stressful family environments (as indexed by lower socioeconomic
status [SES], marital dissatisfaction, and larger family size) were linked to higher levels of PDT within a
large representative sample of Canadian children between the ages of 4 and 11 years ([35]).
As parenting stress has been found to be strongly tied to parental psychological health and well‐being
([22]), we focused on two parental variables: parental malaise (physical and psychological well‐being)
and emotional anger and examined their relationship to PDT. Specifically, prolonged feelings of sadness
and despair, loss of appetite and enjoyment, lethargy, and thoughts about suicide are indicative of
psychological mental state problems. These symptoms, although not always meriting a diagnosis of
psychopathology per se, remain influential in parenting stress and coping processes ([30]). We thus
hypothesized that those mothers reporting high levels of malaise would also be more likely to treat their
children differently. Furthermore, a parent's proneness to emotional anger provides insight into his/her
neuroticism and agreeableness—factors consistently linked to parent–child relationship quality (see [ 5],
for a review). In addition, emotional anger is related to stress ([20]), further underpinning this aspect of
parental temperament as a suitable candidate for risk of PDT.
We also examined three contextual variables: marital satisfaction; household chaos, characterized by a
lack of routines and organization; and parental education as an indicator of SES. These variables were
chosen as they can limit parenting resources ([22], [67]) and are related to parental stress.
As mothers, fathers, and siblings do not interact in isolation, but function within a family system,
dysfunction in one of the subsystems that can be reflected in other subsystems ([49]). Marital
dissatisfaction has been related to lower parental resources, more negative parent–child interaction, and
harsher child discipline (e.g., [28]). When parents are not satisfied in their marital relationship, they may
have fewer personal resources and may not support each other's parenting. This, in turn, can negatively
affect the parent–child relationship and may constrain parental efforts to treat children similarly ([17]).
Previous research has found lower levels of PDT when parents reported elevated levels of marital
communication ([19]) and of marital satisfaction ([35]). Moreover, [57] proposed that conflictual marriages
set the stage for PDT. That is, when mothers and fathers fail to resolve their conflicts, they may misuse
their relationships with their children as part of their conflict ([57]). This idea can be seen when parents
create an alliance with one child, excluding others, thereby increasing differential treatment ([36], [64]).
A more global, though proximal contextual factor also considered was household chaos. A "chaotic
home" is defined as one high in noise and crowding and low in regularity and routines ([66]). Household
chaos has been related to poorer parent–child relationships ([14]), parental stress (e.g., [51]), and
children's behavior problems ([24]). Finally, SES was examined as indicated by parental education.
Substantial literature has shown that parents of lower SES display harsher discipline and more negativity
([16]) in their parent–child relationships ([23]). In addition, parents of lower SES experience elevated
levels of stress and display higher levels of PDT ([17], [35]).
Parenting and Family Type
As well as traditional two‐parent families, the current study included single‐mother families. Reviews of
studies comparing children from divorced versus intact families report that children of divorced parents
exhibit increased levels of behavior problems and that these differences are largely mediated by
differences in parenting in two‐versus one‐parent families (e.g., [ 2]). In other words, parenting in single‐
mother households tends to be less positive and more negative than parenting in the context of two‐
parent families. We know of only one recent study that has demonstrated increased levels of PDT among
single mothers. The study by [35] found greater degrees of differential negativity among the single
mothers in their Canadian sample. They suggested that the higher levels of stress characterizing single
mothers are reflected in elevated levels of parental negativity and decreased abilities and resources to
intentionally keep childrearing equitable. We sought to replicate this finding across multiple informants.
Furthermore, although it is well known that fathers have an influence on children's development ([39]),
most research concerning differential treatment has focused on mothers (cf. [10], [17], [61], [65]). In the
current study, we aimed to investigate both fathers and mothers, thereby extending knowledge about
fathers' differential treatment (FDT).
Finally, we tested for moderation effects by family type. In particular, we hypothesized that the risk of
mothers' differential treatment (MDT) from maternal factors and/or contextual factors might be
accentuated in the context of single‐mother families. This stems from [59] cumulative model of risk and
resilience, in which risk factors operate in a cumulative manner. Accordingly, single parenthood as well as
the parental and contextual factors may increase risk in an interactive rather than in a simple additive
fashion.
Current Study
Our examination of parent and contextual correlates of PDT is the first of its kind in several ways. First,
the proximal contextual factor of household chaos was examined in relation to PDT. Second, we included
reports of parenting from the children themselves (as young as 4 years of age) as well as considering
both mothers' and fathers' reports of their own parenting. Finally, we examined MDT in both two‐parent
and single‐mother households. Specifically, we hypothesized that (a) parental and contextual variables
would be associated with both MDT and FDT; (b) parental and contextual variables would be statistically
predictive of PDT after accounting for the sibling characteristics of differential sibling temperament, sibling
age difference, and sex constellation; and (c) the effects of parental and contextual variables on MDT
would be moderated by family type. Specifically, the risk of MDT from maternal factors and/or contextual
factors would be accentuated in the context of single‐mother families.
Method
Sample and Recruitment
Schools in the south of England were approached and asked to send letters to parents of children in
reception (4‐ to 5‐year‐olds) and Year 1 (5‐ to 6‐year‐olds) classes who had an older brother or sister age
8 years or younger. However, many were unable (or unwilling) to target specific children and sent letters
to all children in these classes. In addition, letters were sent home via the children; therefore, there was
no guarantee that parents received our letters. Because of this opt‐in procedure, it was not possible to
estimate refusal rates accurately.
The sample consisted of 172 families with at least two children. Eighty‐three percent of the older children
in the study were the eldest child in their family, and 79% of the younger children in the study were also
the youngest in their family. Sixty‐three percent of the families who took part had two children, 31% had
three, and 6% had four or more children. Fifty‐four were single‐mother families and the remaining 118
were two‐parent families. Within the two‐parent families, approximately equal numbers of the four sibling
sex constellations (boy–boy, boy–girl, girl–girl, and girl–boy) were recruited, but in the single‐mother
families, boy–girl pairs were overrepresented. Overall, 39 boy–boy pairs, 52 boy–girl pairs, 41 girl–girl
pairs, and 41 girl–boy pairs took part. The average age of the older sibling taking part in the study was
7.4 (SD= 0.84) years, and the average age of the younger sibling was 5.2 (SD= 0.61) years. There were
no significant differences between the mean children's ages across family type. Single mothers (M=
34.19 years, SD= 5.60) were significantly younger than the mothers in the two‐parent families (M= 37.17
years, SD= 4.40). Families came from a mix of working‐class and middle‐class backgrounds, and there
was a wide range of educational attainment among the families. However, single mothers left education
at a significantly earlier stage than mothers within two‐parent families. The families were almost
exclusively Caucasian (92% of mothers), reflecting the population from which this sample was drawn.
Procedure and Measures
Families were visited at home where parents and children were interviewed and parents completed
questionnaires. A brief description of the content of the measures is given next.
Parenting
Child reports The Berkeley Puppet Interview (BPI; [ 1]) is a technique that obtains questionnaire‐type
data from young children using interview questions from two puppets. During the audiotaped interview,
two identical puppets make opposing statements about a member of their family (e.g., "My mom is nice to
me"; "My mom is not nice to me") and then ask the child about themselves (e.g., "How about your
mom?"). Children's responses were subsequently coded on a 7‐point scale where 1 is the most negative
score and 7 the most positive. When a child chooses a response option as expressed by the puppet, a
Code 2 (for a negative response—"My mom is not nice to me") or a Code 6 (for a positive response—"My
mom is nice to me too") is used. When a child amplifies a statement (e.g., "My mom is horrible to me" or
"My mom is really nice to me"), a Code 1 (negative) or 7 (positive) is used. A Code 3 or 5 indicates a
response that is qualified in some way (e.g., "My mom isn't nice to me most of the time" or "My mom is
nice to me most of the time"). Finally, a Code 4 is used when a child indicates that both response options
apply to them.
The interview was composed of two subscales related to parenting—warmth and hostility. Internal
consistencies for the BPI subscales ranged from αs =.62 to.74. The parent–child relationship subscales
of the BPI each contain six items. The warmth subscale includes items such as "My mom is nice to me"
versus "My mom is not nice to me" and the hostility subscale contains items such as "My mom is mean to
me" versus "My mom is not mean to me." After initial coding, scores were calculated such that higher
scores indicate more hostility and warmth, respectively.
Parent reports In order to create warmth and hostility scales that were as equivalent as possible to the
children's puppet reports, we selected items from the Expression of Affection Inventory ([33]), the Parent–
Child Relationship Scale ([33]), the Parental Feelings Questionnaire ([21]), and the Parental Discipline
Interview ([21]). For each of the items from the children's puppet interview, we selected the item closest in
terms of content. For example, parents were asked how often they shouted at their child (from the Parent
Discipline Interview), which we deemed a match for the puppet item "My mom shouts at me a lot."
Average warmth and hostility scores were thus constructed for mothers' and fathers' reports of their
parenting toward the older and younger siblings. Adequate alphas (ranging from.62 to.79) were achieved,
especially in light of the fact that each subscale contained six items.
Child temperament Parents completed the Emotionality, Activity, and Sociability (EAS) ([13]). For the
purpose of this article, we have used the emotionality scales, which consist of five items (e.g., "Child gets
upset easily"). Parents answered using a 5‐point scale, how characteristic or typical each statement was
for their child, ranging from 1 =not characteristic or typical of my child to 5 =very characteristic or typical
of my child. Internal consistencies for mothers' and fathers' reports were excellent (αs =.85–.86). Mother–
father agreement was also substantial (rs =.54 and.51, p <.001, for older and younger siblings,
respectively) and therefore supported the calculation of average emotionality scores for each child.
Parental Factors
Parental anger The measure of parents' emotional anger is taken from the Emotionality, Activity,
Sociality, and Impulsivity (EASI) temperament survey ([13]). This self‐report measure lists a number of
characteristics that are then rated on a 5‐point scale from 1 (strongly disagree) to 5 (strongly agree). The
emotional anger subscale consists of five items (α=.67, α=.68, for maternal and paternal reports,
accordingly) including "There are many things that annoy me" and "I yell and scream more than most
people my age." To avoid confusion with the child emotionality measure, parents' emotional anger is
referred to as parental anger throughout.
Malaise The Malaise Inventory ([31]) is a 24‐item self‐report questionnaire measuring health and general
well‐being. The items cover both physiological (e.g., "Do you often have back‐ache?") and psychological
(e.g., "Do you often feel miserable or depressed?") states and require a yes–no answer (α=.78, α=.74, for
maternal and paternal reports, accordingly).
Contextual Factors
Marital relationship The Golombok Rust Inventory of Marital State ([58]) was used to index both
mothers' and fathers' overall satisfaction with their marital (or cohabiting) relationship. The questionnaire
contains 28 items and is rated on a 4‐point scale. Partners rate their feelings to each statement (e.g., "My
partner is usually sensitive to and aware of my needs"; "I sometimes feel lonely even when I am with my
partner") on a scale of 1 =strongly disagree to 4 =strongly agree. The measure is scored such that higher
scores indicate more marital dissatisfaction. The scale yielded alphas of.89 for both mothers and fathers
in the current study. Mothers' and fathers' reports of their marital satisfaction were substantially correlated
(r=.57, p <.001); therefore, the two reports were averaged to yield a single marital satisfaction score for
each two‐parent family.
Household chaos The Confusion, Hubbub, and Order Scale (CHAOS; [42]) is a 6‐item questionnaire
measuring the level of calm within a household (α=.74, α=.76, for maternal and paternal reports,
accordingly). Items are rated on a 5‐point scale ranging from 1 (definitely untrue) to 5 (definitely true) and
include "The children have a regular bedtime routine" and "We are usually able to stay on top of things."
Mothers' and fathers' reports were again highly correlated (r=.52, p <.001), thus their reports were
averaged.
SES SES was indexed by parental education. This was ascertained via parental interview and ranged
from 1 (no qualifications) to 6 (postgraduate education). Education rather than occupational status was
used because we wanted to avoid the interplay between family type (single‐mother vs. two‐parent family)
and decisions about taking on employment outside the home. In the case of the two‐parent families,
mothers' and fathers' levels of educational attainment were highly correlated (r=.59, p <.001); thus, an
educational average was used to depict SES for these families.
Results
Preliminary Analyses
Creating Differential Treatment Scores
In order to calculate PDT from the parental reports, we calculated the absolute difference between the
parent–younger sibling scores and the parent–older sibling scores. This was done first for parental
warmth, separately for mothers and fathers, such that a high score indicates more differential warmth.
This calculation was then repeated for the parental hostility scores, again separately for mothers and
fathers. In this case, a high score indicates larger differences in the hostility parents showed to their
children. These scores will henceforth be referred to as parent‐reported differential warmth and
differential hostility, respectively.
Difference scores using the children's reports were calculated in a similar manner. That is, the absolute
difference between the two siblings' ratings of parental warmth was calculated as was the absolute
differences for the children's reports of hostility from their parents. As with the parental reports, separate
difference scores were created for mothers and fathers. These scores will henceforth be referred to as
child‐reported "differential warmth" and "differential hostility," respectively. Descriptive statistics for these
and all other study measures are contained in Table 1.
1 Descriptive Statistics for All Study Measures
M (or %) SD
PDT Variables
Child reports
  Differential maternal warmth 0.74 0.77
  Differential maternal hostility 1.25 0.98
  Differential paternal warmth 0.68 0.75
  Differential paternal hostility 1.10 0.90
Maternal reports
  Differential warmth 0.35 0.43
  Differential hostility 0.40 0.41
Paternal reports
  Differential warmth 0.29 0.34
  Differential hostility 0.34 0.30
Predictor Variables
Child variables
  Age difference between sibs (months)26.41 8.98
  Same‐sex sibling pairs 45%
  Opposite‐sex sibling pairs 55%
  Differential emotionality 0.88 0.68
Parental variables
  Maternal anger 2.78 0.71
  Maternal malaise 4.18 3.47
  Paternal anger 2.55 0.70
  Paternal malaise 2.98 2.81
Contextual variables
  Parental education 3.16 1.47
  Marital satisfaction 24.61 9.66
  CHAOS 2.42 0.54
1 Note. CHAOS = Confusion, Hubbub, and Order Scale; PDT = parental differential treatment.
In addition, we calculated sibling intraclass correlations for the parenting variables as an alternate, more
intuitive means of assessing the degree of differential treatment reported by the parents and the children.
For example, mothers' reports of warmth toward their two children were correlated at.60, and their reports
of hostility again substantially correlated at.66. Fathers' reports of parenting revealed even greater levels
of consistency in treatment (rs =.75 and.74, for warmth and hostility, respectively). In the case of the
children's reports of maternal treatment, no consistency in treatment was revealed (rs =.08 and.03, for
warmth and hostility, respectively), and for the children's reports of paternal treatment, the intraclass
correlations indicated modest consistency (rs =.29 and.17, for warmth and hostility, respectively). In
summary, the children's reports yielded substantial estimates of PDT, although substantial consistency in
parental treatment was reported by the parents themselves. Still, parents did not report identical
treatment, validating our quest to identify correlates of the modest to moderate degree of PDT reported
by both children and their parents.
Parent–Child Agreement
It is noteworthy that mothers and children agreed to a moderate degree about parental warmth (rs =.26
for older children and.25 for younger children) and hostility (rs =.33 and.34, for older and younger
siblings, respectively). Agreement between the children and their mothers, however, was negligible to
modest in terms of differential treatment (r=.17 for differential warmth and r=.10 for differential hostility).
Fathers and the older children also yielded some agreement in terms of parental warmth (r=.18) and
especially parental hostility (r=.31). However, no agreement was yielded between the younger children
and their fathers (rs =−.12 and.09, for warmth and hostility, respectively), nor did fathers and their
children agree about the extent of differential treatment displayed by fathers (r=.10 for differential warmth
and r=.12 for differential hostility). These findings indicated that parents and children hold distinct views
regarding differential treatment, warranting separate analyses by reporter.
Differences by Family Type
Finally, before addressing our research questions, we checked for mean differences between fathers,
mothers within two‐parent households, and single mothers for the parental variables and between the
two‐parent and single‐mother families for the contextual variables. The only variable to reveal significant
differences between groups was education. Two‐parent households held higher educational qualifications
(M= 3.36, SD= 1.49) than did the single mothers (M= 2.74, SD= 1.34), t(171) =−2.63, p <.01.
Parental and Contextual Correlates of PDT
In order to explore associations between parental and contextual variables and PDT, Pearson
correlations were conducted (see Table 2). Considering first the sibling dyad characteristics, opposite‐sex
sibling pairs reported more differential hostility from their mothers. Sibling pairs with a larger age
difference reported more differential warmth from their mothers and fathers. However, no significant
associations were found for parental reports of PDT and the siblings' gender constellation or age gap.
Conversely, differential emotionality was associated with higher levels of maternal differential warmth and
hostility as reported by the mothers, although none of the associations with the children's reports of PDT
were significant.
2 Correlations Between All Variables and Absolute PDT
Children's
reports
Parental
reports
Differential warmth Differential
hostility
Differential
warmth
Differential
hostility
Mothers Fathers Mothers Fathers MothersFathersMothersFathers
Sibling Dyad
Characteristics
SS vs. OS pair −.02 .07 .20* .07 .02 −.05 −.06 −.04
Sibling age difference .19* .20* .14 .00 −.04 .07 −.05 .05
Differential
emotionality
.08 −.10 −.06 −.00 .26** .14 .17* .13
Parental Variables
Anger .18* .01 .20* −.13 .24** .06 .15 .16
Malaise .15 .03 .14 −.16 .25** .09 .13 .17
Contextual Variables
Parents' education −.03 −.29** −.08 −.00 −.11 .13 .00 −.00
Marital satisfaction −.07 .04 .04 .11 −.03 .02 .02 .01
CHAOS .12 .30** .12 .23* .23** .08 .13 .05
2 Note. SS = same sex; OS = opposite sex; PDT = parental differential treatment; CHAOS =
Confusion, Hubbub, and Order Scale.
3 * p <.05. p <.01. For the parental variables, a very different pattern emerged for mothers and fathers. No significant correlations were found for fathers. In contrast, parental anger was associated with three of the four PDT variables for mothers. Specifically, maternal anger was correlated with differential hostility as reported by the children and differential warmth as reported by both the children and mothers. In order to examine whether these correlations were significantly different for fathers and mothers, r to z transformations were calculated. This revealed that the correlations between maternal and paternal anger and children's reports of differential hostility differ significantly (z=−2.40, p <.01). In addition, the difference in correlations between maternal and paternal anger and child as well as parent reports of differential warmth neared significance (z=−1.26, p=.10; z=−1.40, p=.08, accordingly). Furthermore, malaise was significantly correlated with maternal differential warmth (mothers' reports) but not with paternal warmth (z=−1.25, p=.10). Finally, for the contextual variables, household chaos proved more important than marital satisfaction and education. In fact, no significant associations emerged for marital satisfaction, and only one correlation (with child‐reported differential warmth) emerged for paternal education but not for maternal education (z=−2.03, p <.05). Household chaos followed a similar pattern in that child‐ reported differential fathering emerged as significant correlates but not the children's reports of differential mothering. These differences were close to significance for children's reports of differential warmth (z= 1.40, p=.08) but were not significant for differential hostility (z=−0.82, ns). In addition, household chaos was linked to maternal differential warmth (parent report) but was not significantly linked to paternal differential warmth (parent report). However, the difference between these correlations was not significant (z=−1.19, ns). Statistical Prediction of PDT We conducted a series of hierarchical regression analyses in order to test whether the parental and contextual variables were predictive of PDT after accounting for characteristics of the sibling dyad. Due to multicollinearity concerns, we first assessed the intercorrelations among the predictor variables. In the case of predictors of MDT, 6 of the 28 correlations were significant, and these ranged from.17 to.39 in magnitude. In the case of predictors of FDT, 7 of the 28 correlations were significant, and these ranged from.20 to.31 in magnitude. These negligible to moderate associations justified considering the variables as separate predictors. First, we consider the hierarchical regression analyses predicting FDT because differing family types did not complicate these analyses. In the first step, the three sibling characteristic variables (same‐sex vs. opposite‐sex siblings pairs, age difference, and differential emotionality) were entered in order to control for sibling dyad characteristics. In the second step, we entered the parental and contextual factors. For example, in the prediction of fathers' differential warmth as reported by the children (see Table 3), the second step contained parental emotional anger, parental malaise, education, marital satisfaction, and household chaos. In this case, only education provided independent prediction of moderate effect size (partial r=.27). Looking at Table 3, only household chaos provided moderate independent prediction (partial r=.28) of differential hostility (child reports). Mirroring the zero‐order correlations, no significant predictors were uncovered in the case of fathers' reports of PDT. Therefore, these analyses are not depicted. 3 Hierarchical Regression Analyses Predicting Paternal Differential Warmth and Hostility (Children's Puppet Reports) Variable Paternal differential warmtha Paternal differential hostilityb B SEB β B SEB β Step 1 (sibling dyad characteristics) SS vs. OS pair 0.12 0.16 .08 0.12 0.19.07 Age difference between sibs 0.02 0.01 .19 −0.000.01−.04 Differential emotionality (paternal report) −0.02 0.13 −.02 −0.040.15−.03 Step 2 (parental and contextual factors) SS vs. OS pair 0.13 0.16 .08 0.12 0.19.07 Age difference between sibs 0.02 0.01 .24 −0.000.01−.04 Differential emotionality (paternal report) −0.05 0.13 −.04 −0.080.16−.06 Parental emotional anger −0.05 0.11 −.04 −0.090.14−.07 Parental malaise −0.01 0.03 −.04 −0.040.04−.15 Parental education −0.15 0.06 −.300.02 0.07.04 Marital satisfaction 0.01 0.01 .07 0.00 0.01.02 CHAOS (paternal report) 0.26 0.19 .17 0.49 0.23.28 4 Note. SS = same sex; OS = opposite sex; CHAOS = Confusion, Hubbub, and Order Scale. 5 Total R =.19; R =.04, for Step 1 (ns); ΔR =.15, for Step 2. a 2 2 2 6 b Total R =.10; R =.01, for Step 1 (ns); ΔR =.10, for Step 2 (ns). 7 * p <.05. The analyses were somewhat different for MDT. In the first step, family type (two parent vs. single mother) was entered. In the second step, the three sibling dyad characteristic variables were entered. In the third step, we entered the maternal and contextual variables, excluding marital satisfaction as this yielded no significant zero‐order correlations nor was it relevant for the single‐mother families. The fourth step (containing interaction terms) is considered in the next section, Moderation by Family Type. These four hierarchical regressions are contained in Tables 4 and 5. Because education differed across family type, we also conducted the regression analyses for MDT using residualized scores with the effects of education removed. This resulted in only minor changes (available from the authors); the beta weights for parental and contextual variables remained identical. As can be seen in Table 4, family type was predictive of differential warmth (mothers' reports), though the effect size was modest (partial r=−.16), indicating that the single mothers were more likely to report treating their children differently than were mothers in two‐parent families. In only one case (see Table 4) did one of the parental/contextual variables provide independent prediction (as shown in Step 3) after controlling for sibling dyad characteristics (Step 2). Maternal anger provided modest independent prediction (partial r=.18) of differential warmth (mothers' reports). The remaining effects of the parental and contextual factors were moderated by family type, as described below. 4 Hierarchical Regression Analyses Predicting Maternal Differential Warmth and Hostility (Mothers' Reports) Variable Maternal differential warmtha Maternal differential hostilityb B SEB β B SEB β Step 1 Family type −0.15 0.08 −.16* −0.010.07−.02 Step 2 Family type −0.16 0.07 −.17* −0.020.07−.02 SS vs. OS pair −0.30 0.70 −.03 −0.070.07−.08 Age difference between sibs −0.00 0.00 −.04 −0.000.00−.03 Differential emotionality 0.17 0.05 .26 0.10 0.05.17*
Step 3
Family type −0.15 0.07 −.16* −0.030.07−.03
SS vs. OS pair −0.03 0.07 −.03 −0.070.07−.09
Age difference between sibs −0.00 0.00 −.04 −0.000.00−.04
Differential emotionality 0.14 0.05 .23** 0.09 0.05.15
Parental emotional anger 0.12 0.05 .19* 0.07 0.05.12
Parental malaise 0.01 0.01 .09 0.01 0.01.04
Education −0.03 0.02 −.09 0.01 0.02.02
CHAOS 0.04 0.07 .05 0.03 0.07.04
2 2 2
Step 4
Family type −0.14 0.07 −.15 −0.030.08−.04
SS vs. OS pair −0.03 0.07 −.03 −0.070.07−.08
Age difference between sibs −0.00 0.00 −.03 −0.000.00−.05
Differential emotionality 0.13 0.05 .20* 0.09 0.05.15
Parental emotional anger 0.15 0.05 .24** 0.03 0.06.13
Parental malaise 0.01 0.01 .11 0.07 0.01.02
Education −0.04 0.03 −.14 0.00 0.08.03
CHAOS 0.07 0.07 .08 0.03 0.08.04
Parental Emotional Anger × Family
Type
−0.14 0.05 −.23**−0.010.06−.01
Parental Malaise × Family Type −0.01 0.01 −.08 0.01 0.01.049
Education × Family Type 0.05 0.03 .15 −0.000.03−.002
CHAOS (maternal report) × Family
Type
−0.00 0.07 −.00 0.02 0.08.029
8 Note. SS = same sex; OS = opposite sex; CHAOS = Confusion, Hubbub, and Order Scale.
9 Total R =.27; R =.03, for Step 1 (p <.05); ΔR =.07, for Step 2 (p <.05); ΔR =.08, for Step 3 (p <.05);
ΔR =.09, for Step 4 (p <.01).
10 b Total R =.07; R =.00, for Step 1 (ns); ΔR =.04, for Step 2 (ns); ΔR =.03, for Step 3 (ns); ΔR =.00,
for Step 4 (ns).
11 * p <.05. *p <.01. 5 Hierarchical Regression Analyses Predicting Maternal Warmth and Hostility (Children's Puppet Reports) Variable Maternal differential warmtha Maternal differential hostilityb B SEB β B SEB β Step 1 Family type −0.15 0.14 −.09−0.160.19−.08 Step 2 Family type −0.13 0.14 −.08−0.120.18−.05 SS vs. OS pair −0.01 0.13 −.000.41 0.17.21
Age difference between sibs 0.02 0.01 .20* 0.01 0.01.13
Differential emotionality 0.10 0.09 .09 −0.100.12−.07
Step 3
Family type −0.14 0.14 −.09−0.090.18−.04
SS vs. OS pair −0.02 0.13 −.020.40 0.17.20*
Age difference between sibs 0.02 0.01 .19* 0.01 0.01.13
Differential emotionality 0.07 0.10 .07 −0.140.13−.02
Parental emotional anger 0.15 0.10 .14 0.19 0.13.13
a 2 2 2 2
2
2 2 2 2 2
Parental malaise 0.02 0.02 .07 0.01 0.03.05
Education 0.01 0.05 .02 −0.060.06−.08
CHAOS 0.07 0.15 .05 0.13 0.18.07
Step 4
Family type −0.15 0.14 −.09−0.060.18−.03
SS vs. OS pair −0.05 0.13 −.030.37 0.12.18*
Age difference between sibs 0.02 0.01 .22* 0.02 0.01.16
Differential emotionality 0.07 0.10 .07 −0.170.30−.12
Parental emotional anger 0.23 0.11 .21* 0.23 0.50.17
Parental malaise 0.02 0.03 .66 0.04 0.03.14
Education 0.02 0.05 .04 −0.040.06−.06
CHAOS 0.09 0.15 .06 0.10 0.19.06
Parental Emotional Anger × Family
Type
−0.21 0.11 −.20−0.310.14−.23*
Parental Malaise × Family Type −0.00 0.03 −.01−0.040.03−.15
Education × Family Type −0.02 0.05 −.030.02 0.07.04
CHAOS (maternal report) × Family
Type
0.03 0.15 .02 0.39 0.18.21*
12 Note. SS = same sex; OS = opposite sex; CHAOS = Confusion, Hubbub, and Order Scale.
13 Total R =.13; R =.01, for Step 1 (ns); ΔR =.05, for Step 2 (ns); ΔR =.04, for Step 3 (ns); ΔR =.03,
for Step 4 (ns).
14 b Total R =.19; R =.01, for Step 1 (ns); ΔR =.06, for Step 2 (p <.05); ΔR =.05, for Step 3 (ns);
ΔR =.08, for Step 4 (p <.05).
15 * p <.05.
Moderation by Family Type
In order to test whether the effects of parental or contextual variables on MDT differed for single mothers
versus those in two‐parent families, we included interaction variables, Parental Variable × Family Type
and Contextual Variable × Family Type in a final step of the regression analyses. All interaction terms
were created using centered data. In the case of differential warmth (mothers' reports) and differential
hostility (children's reports), the Parental Anger × Family Type interactions were significant (see Tables 4
and 5). This interaction also showed a trend in the case of differential warmth according to the children's
puppet reports (Table 5). Although small in magnitude (partial rs =−.17 to −.22), these results indicate that
the pattern of statistical prediction of MDT from maternal anger did differ for the two family types. The
zero‐order correlations indicated substantial links for the single‐mother families (r=.44, r=.53, r=.42, p
<.001, accordingly) and negligible links for the two‐parent families (.08,.07,.09, ns, accordingly). In order
to further examine the nature of these interactions, we first divided the sample into "low maternal anger"
and "high maternal anger" groups using a median split. We then examined the mean MDT scores for
each group by family type. The pattern was clear—differential maternal treatment was more pronounced
for single mothers reporting high emotionality (see Figures 1–3). In the case of differential hostility
a 2 2 2 2 2
2 2 2 2
2
(children's reports), CHAOS × Family Type also provided significant, though modest, independent
prediction (partial r=−.19; see Table 5). In this case, the zero‐order correlations indicate that CHAOS is
linked to MDT in two‐parent but not in single‐mother families (r=.18, p <.05; r=−.03, ns, accordingly). An
inspection of the means revealed the least MDT in low‐CHAOS two‐parent families and the most MDT in
high‐CHAOS two‐parent families (see Figure 4).
Graph: 1 Mean levels of mother‐reported differential warmth for low‐ and high‐anger mothers in single‐
mother and two‐parent families.
Graph: 2 Mean levels of child‐reported differential warmth for low‐ and high‐anger mothers in single‐
mother and two‐parent families.
Graph: 3 Mean levels of child‐reported differential hostility for low‐ and high‐anger mothers in single‐
mother and two‐parent families.
Graph: 4 Mean levels of child‐reported differential hostility for low– and high–Confusion, Hubbub, and
Order Scale (CHAOS) households in single‐mother and two‐parent families.
Variance Explained
As well as considering the independent statistical prediction of each individual risk factor, we examined
the total variance explained by these regression models (listed below each table). Because of the limited
collinearity, the predictors explained a moderate to substantial amount of variance en masse, even
though the effect size for any one predictor was usually modest in magnitude. On average, the predictors
explained 15% of the variance in children's reported PDT, and 17% of mother‐reported PDT. We were
less successful in accounting for variance in FDT as reported by the fathers themselves, explaining on
average 6%.
Discussion
The main goal of our study was to move beyond the examination of links between PDT and child
adjustment, as well as shifting from the large interest in the correlates of parenting ([ 6]), to an exploration
of the correlates of differential parenting. We examined parental and contextual correlates for mothers
and fathers within two‐parent families, as well as for single mothers. Discussion of these results, their
implications, limitations, and future directions follow. First, however, we consider our dual‐informant
(parents and children) approach to PDT.
Using the puppet interviews, it was possible to assess children's perceptions of their relationships with
their parents. The overall low level of agreement seen for parents' and children's reports of PDT supports
previous findings suggesting that despite living in the same family, children and parents often develop
distinct perceptions of parental behaviors and family life events ([38], [40]). Furthermore, replicating
Kowal et al. (2006), we found that in the case of differential treatment, it is very common for parents and
children to perceive events in markedly different ways. The modest to moderate degree of parent–child
agreement of differential treatment seen in this study, as well as the different patterns of associations
discussed later, demonstrates the need to consider children's own views of their treatment. In addition,
previous reports indicate that parents overestimate the consistency of their relationships with their
children ([15], [54]). Therefore, children's reports are particularly important.
Replicating previous studies, we found some evidence for links between sibling dyad–specific
characteristics (sibling age gap, sex constellation, and temperamental difference) and PDT (e.g., [10],
[17], [35]). As PDT may be normative and acceptable when based on child and/or sibling dyad
characteristics, these correlates may reflect "justifiable" PDT. For example, children may view being
warmer to a younger child due to their developmental‐level justifiable PDT, thereby ameliorating its
effects ([37]). In the current study, we were interested in capturing deleterious PDT. Therefore, we
statistically controlled for sibling characteristics before examining parent and contextual factors.
Patterns of association were quite different for mothers and fathers. Correlations revealed that maternal
malaise and especially maternal anger were linked with increased MDT, as was expected based on
previous studies (e.g., [25]). However, these factors were unrelated to FDT. Mothers continue to be more
involved in childrearing (as measured in terms of time spent present with children, as well as actively
engaging with children) than are fathers ([39]). In addition, mothers shoulder more of the responsibility for
child‐care tasks (e.g., taking children to and from school, keeping dentist appointments). The fact that the
parenting role is central to mothers' lives clarifies the reason why mothers' personal resources were
closely tied to the relationships mothers have with their children, differential or otherwise. It should be
noted, however, that these differences in the pattern of correlations between mothers and fathers require
replication and should be interpreted with caution, especially as many of the comparisons indicated only
trend‐level differences between correlations.
Contrary to previous reports (e.g., [17], [35]), marital satisfaction was unrelated to PDT in our sample.
Instead, our novel findings suggest that in houses characterized by elevated levels of disorganization,
both parents tend to exhibit more differential treatment. It is particularly noteworthy that these links were
evident for the children's reports that are not affected by shared method variance. In the case of mothers,
the regression analyses appear to suggest that contextual factors are unimportant. However, post hoc
analyses revealed that contextual factors were predictive after accounting for sibling dyad characteristics;
instead, it is the moderate overlap between the maternal variables and the household chaos that
accounts for the lack of significant prediction among mothers (no such overlap was revealed for fathers).
We interpret these results to mean that mothers who are advantaged in terms of their temperament and
mental health foster more organized (less stressful) households. We draw this as the most sensible
conclusion given that households do not organize themselves—people, and primarily adults, are
responsible for homes being calm and organized rather than chaotic in character. This interpretation was
also supported by the fact that a post hoc mediation analysis indicated that household chaos partially
mediated the links between maternal malaise and anger with differential warmth (parent reports; details
available from the first author). It is not surprising that it is the mothers' rather than fathers' personal
resources that are most important for this sample because of the relatively traditional setup of most of
these families. We would not, however, generalize to all families to state that mothers are responsible for
maintaining family organization. It is the case, however, that for this normative sample, a majority of
mothers were the primary caregivers and spent more time working in the home than did their partners.
It is noteworthy that in the case of fathers, only contextual factors provided statistical prediction of
differential treatment over and above sibling dyad characteristics. In particular, household chaos and
education were predictive of the children's reports of differential parenting. This may be because the role
of fathers is less scripted than that of mothers and more open to the influences of external scaffolding
([39]). Recalling that these are links with differential parenting, we interpret this to mean that household
chaos is particularly stressful for fathers. Similarly, lower SES is typically characterized by elevated stress
([17], [35]). Under such stress, fathers "may become less consciously or intentionally equitable and more
driven by preferences or child characteristics in their childrearing efforts" ([32], p. 47).
Human behavior and development are a result of an interaction between the person and the environment
([12]). Previous research has primarily focused on family cohesion or home stimuli (e.g., the HOME) as
measures of the family environment. Using the CHAOS, we demonstrated that household characteristics
such as regularity, routines, noise, and crowding also play an important role for both fathers' and mothers'
parenting behavior. These results stress the importance of the immediate household environment as
opposed to more distal indices such as SES. In addition, it may be the CHAOS measure that serves as a
marker for generalized, and perhaps covert, family strain, in addition to its function in assessing the overt
household environment.
Examining the regression models, a striking finding is the difference in the way FDT was related to the
sibling dyad characteristics as well as to the parental and contextual variables according to informant. Not
a single significant association emerged when using fathers' reports of differential treatment. A possible
explanation is the exceedingly high levels of consistency in treatment reported by the fathers, yielding
little FDT to explain. Although the fathers saw themselves as mostly consistent in the way they treated
their children, their children disagreed. This is reflected not only in the children's reports of FDT but also
in the fact that a moderate degree of their reports was explained by contextual factors. Findings such as
these highlight the fact that children are active constructors of their social environments and emphasize
the importance of considering children's viewpoints about disparities in parental behaviors. Finally, even
though not predicting the same MDT, it is striking that the correlates are similar for children's and
mothers' reports of MDT.
Correlates of MDT did not act in the same manner for single‐ and two‐parent families. Although only one
of the four tests yielded significant mean‐level differences between single‐and two‐parent families in
MDT, when exploring statistical prediction for the two family types, a distinct picture emerged. Single
mothers were at higher risk of treating their children differently only in the presence of high maternal
anger. That is, the cumulative effect of two risk factors (single parenthood and high maternal anger)
resulted in higher MDT. These results were similar when using both mothers' and children's reports of
PDT. This "double‐risk" finding supports the conclusion made by [60], suggesting that risk factors operate
in a cumulative manner.
In addition, an unexpected pattern of results was found when examining the moderation of the links
between contextual factors and MDT by family type. Mothers from two‐parent families showed the lowest
levels of differential hostility only when coupled with low household chaos. Furthermore, they showed
similar levels of MDT as the single mothers in the presence of high household chaos. These results
suggest that as single mothers may have a "double risk" in the presence of high parental risk, mothers
from intact families may have a "double buffer" when in combination with low contextual risk.
Why is the maternal factor more influential for single mothers, whereas the contextual factor is more
influential for mothers from two‐parent families? One possible explanation is that for single mothers, their
own psychological characteristics and resources cannot be buffered or dominated by a partner's
psychological resources. As single mothers with fewer psychological resources are the only resident
carer of the children and often have less emotional support than married mothers ([69]), they may be
prone to particular difficulties when their own resources are stretched. In two‐parent families, however,
mothers with less psychological resources may receive more support from their partners and therefore
reduce the likelihood that it will affect the way they behave to their children. Moreover, low contextual risk
only acted as a buffer for mothers in two‐parent families. These findings support the growing literature on
the vulnerability of single mothers ([ 2]), which indicates that single motherhood acts as a risk factor for
child development and for the mother–child relationship but mainly when coupled with other risk factors
([60], [68]). Conversely, we found that being a mother in a two‐parent family does not act as a protective
factor in and of itself. However, when coupled with a factor such as low environmental chaos, it may act
as a buffer and in fact be linked to lower MDT.
Implications
To date, most research on PDT has focused on child or sibling dyad characteristics assuming that PDT is
predominantly affected by the siblings' age, gender, and temperament (e.g., [10]). In fact, some previous
work has concluded that the vast majority of links between PDT and children's outcomes are due to
differences in the siblings' genes rather than being a pure environmental effect ([53], [63]). However, the
results in this study indicate that parental and contextual factors also contribute to PDT. We would argue
that PDT related to parental and contextual factors may be particularly potent given that it is not merely
reflecting siblings' differing genetic propensities. It is also impressive that some of our findings replicated
across both child and parent reports, and that the parental and contextual factors accounted for nontrivial
amounts of variation in differential parenting (on average 15% in the case of the children's puppet reports
and 17% in the case of maternal reports). Finally, the different pattern of risk for mothers and fathers
highlights the need to examine PDT in a more refined manner that uncovers the unique sources of stress
for different parents.
As we would expect from family systems theory (see [34]) and research (e.g., [41]), our findings
demonstrated that relationships between individuals in the family are part of a larger system. The family
system is affected by contextual factors as well as individual characteristics of family members. Our
findings also echo [70] in pointing to the value of attending to all family members when providing family
support rather than treating each in isolation. Recent policy and practice interest have focused on the
parent–child relationship, at the expense of other family relationships. Our findings suggest that children
are equally as likely as parents to reap the benefits of services or interventions directed toward
enhancing parents' well‐being. Furthermore, although associations were found between stress variables
and PDT for both parents, the specific type of stress and its origin may vary for mothers and fathers. The
different type of stressors linked to MDT and FDT can inform the development of future intervention
programs. Similarly, the "double‐risk" phenomenon seen for the single mothers highlights the importance
of supporting single mothering, reducing stress originating from their personal resources and thus
reducing risk for MDT.
Limitations and Future Research
We only considered two children within a limited age range in each family. Some of the families, however,
had three or more children. This means that the way parents behave to their children and the way
children perceive the differential treatment may be more complex and needs to be further explored
considering all family members. Second, although the current sample included working‐ and middle‐class
families, the population from which the sample was drawn is primarily Caucasian. Recent research by
Brody and colleagues has examined parenting and child outcomes among African American siblings
(e.g., [ 7]). However, we are unaware of research examining the correlates of PDT in ethnic minority
populations. As Britain has become a multicultural country with a diversity of ethnic minority groups ([50]),
it is important to replicate this study with ethnically diverse samples. Future research including ethnic
minority groups and families across different cultures has the potential to increase the generalizability of
current findings or to uncover cultural variations reflecting differing family processes. Third, although one
of the advantages of the current study is using both parental and children's reports of PDT, it did not
include observational measurements of parenting. A future study involving parent–child interactions could
reveal interesting facets depending on whether PDT is objectively or subjectively measured. In addition,
future research examining father involvement may uncover whether differences in correlates of PDT for
fathers and mothers were related to the amount of time fathers spend in the company of their children.
Finally, the current study was conducted at a single time point. Consistency in PDT has been previously
reported ([43]). However, we are unaware of research examining the correlates of PDT over time. A
longitudinal study across different developmental periods would add to our understanding of change and
stability in "determinants" of PDT.
References
1 Ablow, J. C., & Measelle, J. R. (1993). The Berkeley Puppet Interview: Interviewing and coding systems
manuals. Eugene: University of Oregon, Department of Psychology.
2 Amato, P. R. (2001). Children of divorce in the 1990s: An update of the Amato and Keith (1991) meta‐
analysis. Journal of Family Psychology, 15, 355 – 370.
3 Anderson, K. E., Lytton, H., & Romney, D. M. (1986). Mothers' interactions with normal and conduct‐
disordered boys: Who affects whom? Developmental Psychology, 22, 604 – 609.
4 Belsky, J. (1984). The determinants of parenting: A process model. Child Development, 55, 83 – 96.
5 Belsky, J., & Barends, N. (2002). Personality and parenting. In M. H. Bornstein (Ed.), Handbook of
parenting: Vol. 3. Being and becoming a parent (2nd ed., pp. 415 – 438). Mahwah, NJ: Erlbaum.
6 Belsky, J., & Jaffee, S. R. (2006). The multiple determinants of parenting. In D. Cicchetti & D. J. Cohen
(Eds.), Developmental psychopathology: Vol. 3. Risk, disorder, and adaptation (2nd ed., pp. 38 – 85).
Hoboken, NJ: Wiley.
7 Brody, G. H., Kim, S., Murray, V. M., & Brown, A. C. (2005). Longitudinal links among parenting, self‐
presentation to peers, and the development of externalizing and internalizing symptoms in African
American siblings. Development and Psychopathology, 17, 185 – 205.
8 Brody, G. H., & Stoneman, Z. (1990). Sibling relationships. In G. H. Brody & I. E. Sigel (Eds.), Methods
of family research: Biographies of research projects: Vol. 1. Normal families (pp. 189 – 212). Hillsdale,
NJ: Erlbaum.
9 Brody, G. H., Stoneman, Z., & Burke, M. (1987). Child temperaments, maternal differential behavior,
and sibling relationships. Developmental Psychology, 23, 354 – 362.
Brody, G. H., Stoneman, Z., & McCoy, J. K. (1992). Parental differential treatment of siblings and sibling
differences in negative emotionality. Journal of Marriage and the Family, 54, 643 – 651.
Brody, L. R., Copeland, A. P., Sutton, L. S., Richards, D. R., & Guyer, M. 1998). Mommy and daddy like
you best: Perceived family favouritism in relation to affect, adjustment and family process. Journal of
Family Therapy, 20, 269 – 291.
Bronfenbrenner, U. (1979). The ecology of human development: Experiments by nature and design.
Cambridge, MA: Harvard University Press.
Buss, A. H., & Plomin, R. (1984). Temperament: Early developing personality traits. Hillsdale, NJ:
Erlbaum.
Coldwell, J., Pike, A., & Dunn, J. (2006). Household chaos—Links with parenting and child behaviour.
Journal of Child Psychology and Psychiatry, 47, 1116 – 1122.
Coldwell, J., Pike, A., & Dunn, J. (in press). Maternal differential treatment and child adjustment: A multi‐
informant approach. Social Development.
Conger, R. D., Conger, K. L., Elder, G. H., Lornez, F. O., Simons, R. L., & Whitbeck, L. B. (1992). A family
process model of economic hardship and adjustment of early adolescent boys. Child Development, 63,
526 – 554.
Crouter, A. C., McHale, S. M., & Jenkins‐Tucker, C. (1999). Does stress exacerbate parental differential
treatment of siblings? A pattern‐analytic approach. Journal of Family Psychology, 13, 286 – 299.
Daniels, D., & Plomin, R. (1985). Differential experience of siblings in the same family. Developmental
Psychology, 21, 747 – 760.
Deal, J. E. (1996). Marital conflict and differential treatment of siblings. Family Process, 35, 333 – 346.
Deater‐Deckard, K. (1998). Parenting stress and child adjustment: Some old hypotheses and new
questions. Clinical Psychology: Science and Practice, 5, 314 – 332.
Deater‐Deckard, K. (2000). Parenting and child behavioral adjustment in early childhood: A quantitative
genetic approach to studying family processes. Child Development, 71, 468 – 484.
Deater‐Deckard, K. (2005). Parenting stress and children's development: Introduction to the special
issue. Infant and Child Development, 14, 111 – 115.
Deater‐Deckard, K., Atzaba‐Poria, N., & Pike, A. (2004). Mother‐ and father‐child mutuality in Anglo and
Indian British families: A link with lower externalizing problems. Journal of Abnormal Child Psychology,
32, 609 – 620.
Dumas, J. E., Nissley, J., Nordstrom, A., Smith, E. P., Prinz, R. J., & Levine, D. W. (2005). Home chaos:
Sociodemographic, parenting, interactional, and child correlates. Journal of Clinical Child and Adolescent
Psychology, 34, 93 – 104.
Dunn, J., & Plomin, R. (1986). Determinants of maternal behavior towards 3‐year‐old siblings. British
Journal of Developmental Psychology, 4, 127 – 137.
Dunn, J., & Plomin, R. (1990). Separate lives: Why siblings are so different. New York: Basic Books.
Dunn, J., Stocker, C., & Plomin, R. (1990). Nonshared experiences within the family: Correlates of
behavioral problems in middle childhood. Development and Psychopathology, 2, 113 – 126.
Erel, O., & Burman, B. (1995). Interrelatedness of marital relations and parent‐child relations: A meta‐
analytic review. Psychological Bulletin, 118, 108 – 132.
Feinberg, M., & Hetherington, E. M. (2001). Differential parenting as a within‐family variable. Journal of
Family Psychology, 15, 22 – 37.
Goodman, S. H., & Gotlib, I. H. (Eds.). (2002). Children of depressed parents: Mechanisms of risk and
implications for treatment. Washington, DC: American Psychological Association.
Grant, G., Nolan, M., & Ellis, N. (1990). A reappraisal of the Malaise Inventory. Social Psychiatry and
Psychiatric Epidemiology, 25, 170 – 178.
Henderson, S. H., Hetherington, E., Mekos, D., & Reiss, D. (1996). Stress, parenting, and adolescent
psychopathology in nondivorced and stepfamilies: A within‐family perspective. In E. A. Blechman & E. M.
Hetherington (Eds.), Stress, coping, and resiliency in children and families (pp. 39 – 66). Mahwah, NJ:
Erlbaum.
Hetherington, E. M., & Clingempeel, W. G. (1992). Coping with marital transitions: A family systems
perspective. Monographs of the Society for Research in Child Development, 57 (Serial No. 227).
Hinde, R. A. (1992). Ethological and relationships approaches. In R. Vasta (Ed.), Six theories of child
development (pp. 251 – 285). London: Jessica Kingsley.
Jenkins, J. M., Rasbash, J., & O'Connor, T. G. (2003). The role of the shared family context in differential
parenting. Developmental Psychology, 39, 99 – 113.
Kitzmann, K. M. (2000). Effects of marital conflict on subsequent triadic family interactions and parenting.
Developmental Psychology, 36, 3 – 13.
Kowal, A., & Kramer, L. (1997). Children's understanding of parental differential treatment. Child
Development, 68, 113 – 126.
Kowal, A. K., Krull, J. L., & Kramer, L. (2006). Shared understanding of parental differential treatment in
families. Social Development, 15, 276 – 295.
Lamb, M. E. (Ed.). (2004). The role of the father in child development (4th ed.). Hoboken, NJ: Wiley.
Larson, R., & Richardson, M. H. (1994). Divergent realities. New York: Basic Books.
Luster, T., & Okagaki, L. (1993). Multiple influences on parenting: Ecological and life‐course perspectives.
In T. Luster & L. Okagaki (Eds.), Parenting: An ecological perspective (pp. 227 – 250). Hillsdale, NJ:
Erlbaum.
Matheny, A. P., Wachs, T. D., Ludwig, J. L., & Phillips, K. (1995). Bringing order out of chaos:
Psychometric characteristics of the confusion, hubbub, and order scale. Journal of Applied
Developmental Psychology, 16, 429 – 444.
McGuire, S., Dunn, J., & Plomin, R. (1995). Mothers' differential treatment of siblings and children's
behavioral problems: A longitudinal study. Development and Psychopathology, 7, 515 – 528.
McHale, S. M., Crouter, A. C., McGuire, S. A., & Updegraff, K. A. (1995). Congruence between mothers'
and fathers' differential treatment of siblings: Links with family relations and children's well‐being. Child
Development, 66, 116 – 128.
McHale, S. M., & Pawletko, T. M. (1992). Differential treatment of siblings in two family contexts. Child
Development, 63, 68 – 81.
McHale, S. M., Updegraff, K. A., Jackson‐Newsom, J., Tucker, C. J., & Crouter, A. C. (2000). When does
parents' differential treatment have negative implications for siblings? Social Development, 9, 149 – 172.
Measelle, J. R., Ablow, J. C., Cowan, P. A., & Cowan, C. P. (1998). Assessing young children's views of
their academic, social, and emotional lives: An evaluation of the self‐perception scales of the Berkeley
Puppet Interview. Child Development, 69, 1556 – 1576.
Mekos, D. (1996, March). Differences in siblings' experience of family hardships. Paper presented at the
Biennial Meeting of the Society for Research in Adolescence, Boston.
Minuchin, S. (1974). Families and family therapy. Cambridge, MA: Harvard University Press.
Modood, T. (1997). Culture and identity. In T. Modood & R. Berthoud (Eds), Ethnic minorities in Britain.
Diversity and disadvantage (pp. 290 – 338). London: Policy Studies Institute.
Ostberg, M., & Hagekull, B. (2000). A structural modeling approach to the understanding of parenting
stress. Journal of Clinical Child Psychology, 29, 615 – 625.
Pike, A., Coldwell, J., & Dunn, J.(2006). Family relationships in middle childhood. York, UK: York
Publishing Services/Joseph Rowntree Foundation.
Pike, A., McGuire, S., Hetherington, E. M., Reiss, D., & Plomin, R. (1996). Family environment and
adolescent depressive symptoms and antisocial behavior: A multivariate genetic analysis. Developmental
Psychology, 32, 590 – 603.
Pike, A., Reiss, D., Hetherington, E. M., & Plomin, R. (1996). Using MZ differences in the search for
nonshared environmental effects. Journal of Child Psychology and Psychiatry, 37, 695 – 704.
Plomin, R., & Daniels, D. (1987). Why are children in the same family so different from one another?
Behavioral and Brain Sciences, 10, 1 – 60.
Radke‐Yarrow, M. (1998). Children of depressed mothers: From early childhood to maturity. New York:
Cambridge University Press.
Reiss, D., Plomin, R., Hetherington, E. M., Howe, G. W., Rovine, M. J., Tryon, A., et al. (1994). The
separate worlds of teenage siblings: An introduction to the study of nonshared environment and
adolescent development. In E. M. Hetherington, D. Reiss, & R. Plomin (Eds.), Separate social worlds of
siblings: The impact of nonshared environment on development (pp. 63 – 109). Hillsdale, NJ: Erlbaum.
Rust, J., Bennum, I., Crowe, M., & Golombok, S. (1990). The GRIMS: A psychometric instrument for the
assessment of marital discord. Journal of Family Therapy, 12, 45 – 57.
Rutter, M. (1979). Maternal deprivation, 1972‐1978: New findings, new concepts, new approaches. Child
Development, 50, 283 – 305.
Rutter, M., Tizard, J., & Whitmore, K. (1970). Education, health and behavior. London: Longmans.
Stocker, C. (1995). Differences in mothers' and fathers' relationships with siblings: Links with children's
behavior problems. Development and Psychopathology, 7, 499 – 513.
Tarullo, L. B., DeMulder, E. K., Ronsaville, D. S., Brown, E., & Radke‐Yarrow, M. (1995). Maternal
depression and maternal treatment of siblings as predictors of child psychopathology. Developmental
Psychology, 31, 395 – 405.
Turkheimer, E., & Waldron, M. (2000). Nonshared environment: A theoretical, methodological, and
quantitative review. Psychological Bulletin, 126, 78 – 108.
Volling, B. L. (1997). The family correlates of maternal and paternal perceptions of differential treatment
in early childhood. Family Relations: Interdisciplinary Journal of Applied Family Studies, 46, 227 – 236.
Volling, B. L., & Belsky, J. (1992). The contribution of mother‐child and father‐child relationships to the
quality of sibling interaction: A longitudinal study. Child Development, 63, 1209 – 1222.
Wachs, T. D. (2005, April). Environmental chaos and children's development: Expanding the boundaries
of chaos. Paper presented at the Society for Research in Child Development, Atlanta, GA.
Wachs, T. D., & Camli, O. (1991). Do ecological or individual characteristics mediate the influence of the
physical environment upon maternal behavior. Journal of Environmental Psychology, 11, 249 – 264.
Weinraub, M., & Gringlas, M. B. (1995). Single parenthood. In M. H. Bornstein (Ed.), Handbook of
parenting: Vol. 3. Status and social conditions of parenting (pp. 65 – 88). Mahwah, NJ: Erlbaum.
Weinraub, M., Horvath, D. L., & Gringlas, M. B. (2002). Single parenthood. In M. H. Bornstein (Ed.),
Handbook of parenting: Vol. 3. Being and becoming a parent (2nd ed., pp. 109 – 140). Mahwah, NJ:
Erlbaum.
Welsh, E., Buchanan, A., Flouri, E., & Lewis, J. (2004). "Involved" fathering and child well‐being: Fathers'
involvement with secondary school age children. London: National Children's Bureau and Joseph
Rowntree Foundation.