You are on page 1of 25

Copyright © eContent Management Pty Ltd. Journal of Family Studies (2011) 17: 86–109.

Family structure, co-parental


relationship quality, post-separation
paternal involvement and children’s
emotional wellbeing
JENNIFER BAXTER PHD1
Senior Research Fellows, Australian Institute of Family Studies, Melbourne, VIC, Australia

RUTH WESTON MA
Principal Research Fellow, Australian Institute of Family Studies, Melbourne, VIC, Australia

LIXIA QU PHD
Senior Research Fellows, Australian Institute of Family Studies, Melbourne, VIC, Australia

ABSTRACT
It is well documented that children who experience parental divorce are more likely than those in intact
families to experience a range of emotional and behavioural adjustment problems, and to perform less well
academically. However, few studies of the impact of divorce have exclusively considered young children.
This paper takes advantage of a recent Australian child cohort study to examine links between young
children’s emotional wellbeing, the quality of the co-parental relationship, and post-separation paternal
involvement. We found that while children aged 6–7 years living with both parents generally had better
emotional wellbeing than similar aged children living with one parent, inter-parental hostility was an
important factor in explaining young children’s emotional wellbeing. But regardless of family type, children
whose parents had a hostile inter-parental relationship tended to have poorer emotional wellbeing than
children whose parents did not have a non-hostile relationship, as reported by children and their parents.

Key words: children; divorce; wellbeing; parental conflict; family structure; family transitions

A s in other western countries, Australia has


experienced striking changes in many aspects
of family life, including increases in the propor-
dependent children, the proportion headed by a
sole parent has increased (e.g., from 15% in 1986
to 22% in 2006; Australian Bureau of Statistics
tions of couples who are cohabiting outside mar- [ABS], 2001, 2007). Most of these families arise
riage, in the fragility of relationships, and in the from the separation of parents who had been mar-
number of children born outside marriage (Hayes, ried to each other (Hayes et al., 2010).
Weston, Qu, & Gray, 2010). It is not surprising, Each year over the past few decades, around
then, that among all Australian families with 50,000 children under age 18 years have

1
Correspondence to: Dr Jennifer Baxter, Australian Institute of Family Studies, Melbourne, Australia 3000;
tel: +61 3 9214 7888; e-mail: Jennifer.Baxter@aifs.gov.au

86 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

experienced parental divorce, However, given that that high levels of overt inter-parental con-
separating parents include those who were mar- flict increase the risk of children experiencing
ried to each other and do not seek divorce (at least socio-emotional problems. These problems
for several years and possibly not at all), as well as that can be long-lasting (see Amato & Booth,
those who had been living together outside mar- 1997; Sarrazin & Cyr, 2007). It is therefore not
riage the number of children experiencing paren- surprising that the elevated risk of adjustment
tal separation is likely to be considerably greater problems apparent for children whose parents
than this figure. Overall, one in five children aged have separated are frequently linked to their
under age 18 years have a natural parent living experiences of significant pre-separation con-
elsewhere (ABS, 2011). flict (Amato & Booth, 1997; Ambert, 1997).
Prior research has shown that children who At the same time, there is some evidence to
experience parental divorce are more likely than suggest that of children exposed to high levels
those in intact families to experience a range of of inter-parental conflict, those whose parents
emotional and behavioural adjustment problems, remain together tend to indicate poorer adjust-
and to perform less well academically. As adults, ment than those whose parents separate (e.g.,
they are also more likely to divorce and become Jekielek, 1997), while divorce between parents
single parents themselves than those who grew up who displayed low conflict prior to separation
in intact families (see reviews by Amato, 2000; tends to be associated with poorer wellbeing for
Cherlin, Chase-Lansdale, & McRae, 1998). While these children (Morrison & Coiro, 1999). In
differences in emotional and behavioural problems his review of this literature, Amato (2010) con-
between children in divorced and intact families cluded: ‘Taken together, these studies suggest
are typically modest, with much overlap apparent, that the consequences of divorce depend on
the absolute number of children adversely affected whether children are removed from an aversive
by divorce is substantial (Amato, 2000). or a supportive family environment’ (p. 657).
Assessments of links between parental separa- While parental separation may bring about
tion and children’s wellbeing need to take into relief from exposure to very frequent and intense
account the fact that relationship breakdown conflict between parents, it also tends to create
does not occur randomly. That is, factors that a range of other disruptions. Immediate upheav-
predispose parental separation may also be asso- als may include substantial increases in financial
ciated with children’s elevated risk of experienc- difficulties and associated deprivations, changes
ing emotional or behavioural problems. Without in housing and school, enhanced parental dis-
taking some of these background characteristics tress and/or diminished quality of parenting,
into account, the impact of parental separation and substantially reduced time or even total loss
on children will be overstated (Furstenberg & of contact with one parent. Such disruptions can
Kiernan, 2001). For instance, parents’ problem- be very distressing for the children (see Amato,
atic personality traits or mental health problems 2010; Carlson & Corcoran, 2001; Furstenberg
that contribute to separation may be inherited et al. 1987; Kelly & Emery, 2003). In forg-
by children, increasing their chance of displaying ing new lives, parents may re-partner, perhaps
various psycho-social problems (Amato, 2010; see more than once They may find themselves lead-
Pryor & Rodgers, 2001). ing very complex lives which include having to
Regardless of whether separation takes take responsibility for additional step children
place, processes associated with deteriorating or additional children resulting from the new
relationships between parents are also likely to relationship (Amato, 2010; Cherlin, 2008).
impact negatively on the children. Thus stud- Parental separation may therefore be best seen as
ies of intact and separated families suggest a step in the process of relationship breakdown

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 87


Jennifer Baxter, Ruth Weston and Lixia Qu

and readjustment, with the outcomes for chil- inter-parental relationship and parent and child
dren linked to their experiences of the entire characteristics were taken into account. On the
process. This process is likely to entail multiple other hand, fathers with shared care time pro-
transitions such as sequential changes in family vided more favourable assessments of their child’s
structure, which may in turn adversely affect the wellbeing than fathers with a minority of care
children (Amato, 2010).2 time, while those who never saw their child pro-
When questioned about their feelings about vided the least favourable assessments.
their parents’ separation, many children express Findings such as those reported above need to
sadness about spending diminished or no time be qualified by the fact that research has consis-
with one parent (usually the father; Lodge & tently shown that different informants are quite
Alexander, 2010; Pryor & Rodgers, 2001). Yet often not in close agreement when assessing
previous research into the links between frequency children’s psycho-social wellbeing (Achenbach,
or duration of post-separation parent–child con- McConaughy, & Howell, 1987; De Los
tact and children’s wellbeing has yielded mixed Reyes & Kazdin, 2005; Loeber, Green, Lahey, &
results.3 On the basis of a meta-analysis of 12 Stouthamer-Loeber, 1991). This is particularly
studies, Whiteside and Becker (2000) concluded the case when informants are reporting on inter-
that the amount of time with the father was part nalising problems, such as anxiety and depression
of the complex model explaining children’s out- (Achenbach et al., 1987; Choudhury, Pimentel, &
comes, with the link between fathers’ time with Kendall, 2003; Comer & Kendall, 2004; Duhig,
children and children’s outcomes being mediated Renk, Epstein, & Phares, 2000; Karver, 2006).
by the quality of the father-child involvement. In addition, individual informants’ assessments
Based on another meta-analysis of 63 studies, can be affected by ‘situation specificity’. For exam-
Amato and Gilbreth (1999) concluded that chil- ple, children may be unhappy or anxious at school
dren’s frequency of contact with their non-resi- but not at home or vice versa, leading teachers
dent parent does not necessarily benefit children. and parents to provide quite different views of the
Rather, the quality of parenting displayed during child’s psycho-social wellbeing. Smart, Sanson,
this time together affects child wellbeing. These Baxter, Edwards, & Hayes (2008) provide an anal-
authors also found that the relationship between yses of child outcomes using teachers’ responses.
paternal contact and child wellbeing was stronger Parents may better reflect how children are at
for recent than earlier studies. They considered home. For example children at home are likely to
this to be a tentative finding which, if true, may be more directly exposed to any parental difficul-
arise from greater commitment and parenting ties that exist or to the realities of a parental sepa-
skills among more recent cohorts of non-resident ration (Emery, 1982; Karver, 2006). But parents’
fathers. reports of their children’s emotional wellbeing may
Kaspiew et al. (2009) examined the link be more directly coloured by their own sense of
between child wellbeing (as reported by one wellbeing (i.e., ‘shared method bias’). Being linked
parent – the mother or the father) and different to parents’ own coping mechanisms and emotional
care-time arrangements of Australian children states, parental perceptions of the impact of family
whose parents had been separated for an aver- problems on their children may be positively or
age of 15 months. Mothers’ assessments of child negatively skewed. For example, there is evidence
wellbeing were not linked with care-time arrange- that mothers with anxiety problems report more
ments, once their views of the quality of the anxiety problems in their children (Briggs-Gowan,
2
At the same time, parental separation is by no means an inevitable consequence of relationship breakdown.
3
It is unusual for studies to examine both the frequency of time spent with the children and typical duration of
such episodes. Smyth (2004) represents an exception.

88 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

Carter, & Schwab-Stone, 1996; Chi & Hinshaw, population. The LSAC study is based on two
2002; Richters, 1992). On the other hand, deal- birth cohorts of children and their families: the
ing with and being distracted by their own prob- younger cohort was born between March 2003
lems, some parents may overlook the emotional and February 2004, and the elder between March
and behavioural difficulties experienced by their 1999 and February 2000. A random sample of
children during and after parental separation. Or children was selected by the HIC and their fami-
regardless of the home situation, parents may wish lies were invited to participate in the study. The
to present their children in the most positive light final sample, comprising 54% of those selected,
possible (Karver, 2006; Riley, 2004; Silverman was broadly representative of all Australian chil-
& Rabian, 1995). On the other hand, teachers dren, with the LSAC sample distribution simi-
are unlikely to know a child as well as the child’s lar to the population distribution (based on the
parents. Australian Census) on a number of characteristics
Gathering information from young children (Australian Institute of Family Studies, 2005).
about their own wellbeing is difficult within the In this article we focus on the elder of these
context of large-scale surveys. More commonly, two cohorts (aged 4–5 years in the first survey
such research, is conducted in clinical settings. wave). The first survey wave was held in 2004
Such assessments are based on the premise that (N = 4,976 for the 4- to 5-year cohort) and the
children are reliable informants of their own feel- second in 2006, when the elder cohort of children
ings and behaviours. While children lack the was 6–7 years old (N = 4,464, 90% of those inter-
sophistication and experience that adults can bring viewed at Wave 1). Wave 2 data have been used in
to judgements of particular emotions and, like this analysis because the children were also inter-
adults, may be influenced by social desirability in viewed at this time. (For a detailed description of
the responses they give (Cole, Hoffman, Tram, the LSAC study design, see Gray & Smart, 2008,
& Maxwell, 2000; Karver, 2006; Riley, 2004; and Soloff, Lawrence, & Johnstone, 2005.)
Silverman & Rabian, 1995), their feelings can nev- The following analyses are based on a sample
ertheless be captured with age-appropriate instru- of 4,303 children. Of these children, 3,640 (85%)
ments (Measelle, Ablow, Cowan, & Cowan, 1998). were living with both biological parents, and 663
In the study reported on below, we explore the (15%) were living with their mother, while their
links between the emotional wellbeing of children father lived elsewhere. In total, 560 were living with
aged 6–7 years, the quality of their parents’ co- their single mother (i.e., 13% of all children rep-
parental relationship and post-separation paternal resented in the present study), and the remaining
involvement. 103 (2%) were living with their mother and step-
father.4 Final numbers for some analyses vary from
DATA AND METHOD these because of missing data for the measures of
The present analyses are based on data from the emotional wellbeing or the explanatory variables.
Growing up in Australia: the Longitudinal Study Several variables were derived to characterise the
of Australian Children (LSAC), an ongoing lon- co-parental relationship. In the present analyses, a
gitudinal study of children’s growth and devel- measure of whether or not the parents believed that
opment. The sampling frame for LSAC was the the relationship entailed hostility at least some of
Health Insurance Commission’s (HIC) Medicare the time was derived. Where the father lived else-
database, a comprehensive database of Australia’s where, this measure was based on a single question

4
A total of 169 children were excluded from analyses for the following reasons: key data about the child’s living
arrangements were missing; the child’s was not living with the mother (there were too few children living with
their single or re-partnered fathers to derive reliable estimates of the children’s wellbeing); the child had a parent
who had either died, was in jail, or living elsewhere because of an apparent ‘temporary’ separation.

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 89


Jennifer Baxter, Ruth Weston and Lixia Qu

asked of mothers only: ‘How often is there anger which the child has worries or fears, is unhappy,
or hostility between you?’. Mothers who indicated and/or is nervous or clingy. Assessments by moth-
that anger or hostility sometimes, often, or always/ ers were available for 4,189 children (97% of the
almost always prevailed were deemed to have a in-scope sample). Teacher assessments were pro-
‘hostile’ inter-parental relationship, while those vided for 3,487 children (81% of the sample).
who said that anger or hostility was never/almost The emotional symptoms scale ranges from 0
never or rarely were considered to have a ‘non-hos- to 10, with a higher number corresponding to a
tile’ relationship. For intact couples, the co-paren- greater number of negative emotional symptoms.
tal relationship was classified as hostile or not on In Wave 2, the children were asked a set of
the basis of parents’ answers to the following two questions about their enjoyment of school, their
questions: (a) ‘How often is there anger or hostility relationships with teachers and other students,
between you?’; (b) ‘How often do you have argu- and also about their feelings. The latter ques-
ments with your partner that end up with people tions were used in this analysis and constitute the
pushing, hitting, kicking or shoving?’ Response third measure of emotional wellbeing used in our
options were: never, rarely, sometimes, often, or analysis. Specifically, the children were asked to
always. Relationships were classified as ‘hostile’ if indicate how often they felt (a) happy, (b) scared
a parent reported that either one of these experi- or worried, (c) sad or (d) angry or mad. Possible
ences occurred at least ‘sometimes’.5 This approach response categories were: lots of times (= 1), some-
was based on the assumption that parents would times (= 2) and hardly ever (= 3). They were also
be inclined to minimise, or feel reluctant to dis- shown prompt cards with these categories written
close, relationship difficulties. on them, and with each category paired with a
All other measures in this set of analyses were simple drawing of a face showing a smile, appar-
based on mothers’ reports (a) because very little ent indifference, or a frown. In analysing these
information was derived from parents who lived data, the coding for the last three items was
elsewhere (all of whom were fathers in the present reversed. The scores for each item were summated
analyses), and (b) because not all fathers in intact (range = 4–12) and then rescaled so that, as for
families provided information.6 the SDQ Emotional Difficulties scale, total scores
A unique feature of the LSAC study is the ranged from 0–10, with a higher score depicting
availability of multiple perspectives on children’s greater evidence of distress. Cronbach’s alpha was
wellbeing: parents’ reports (usually the mother), quite low for the scale derived from these items
teachers’ reports and the reports of children (α = 0.42). Accordingly, multivariate analyses
themselves. Three measures of children’s experi- were also conducted separately on each of the
ence of emotional wellbeing were thus developed, underlying items to determine whether results
two of which were based on mothers’ and teach- reflected certain items (results not shown). The
ers’ answers to the multi-item ‘emotional symp- associations presented in this paper, for the overall
toms’ subscale of the Strengths and Difficulties scale, usually reflect associations apparent for each
Questionnaire (SDQ: Goodman, 2001). This of the underlying items, although in the separate
subscale captures, for example, the extent to analyses the effect sizes tended to be greatest for

5
If hostility is instead grouped such that ‘often’ and ‘always’ are compared to ‘sometimes’, ‘rarely’ or ‘never’, then
many significant differences according to this variable are no longer apparent. This may be due to the treatment
of ‘sometimes’ as the same as those with less possible exposure to conflict, or it may be due to the smaller
number classified as hostile, reducing the ability to detect significant differences. Using this definition, across all
single and couple families, 202 are classified as having some hostility compared to 1,021 using the more inclusive
definition.
6
A questionnaire designed for parents who lived elsewhere was available in Wave 2, but attracted a very low
response rate.

90 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

‘angry or mad’. These questions were answered by 4. Children who were living with one parent,
4,220 children (98% of children). where mothers did not report a hostile rela-
It is important to note that, while the three tionship with the father; and
measures described above are actually measures 5. Children who were living with one parent,
of emotional difficulties, we refer to the findings and never spent any time with their father
throughout from the perspective of wellbeing.
We do this for two reasons: (a) we seek to avoid For children of separated parents, we also incorpo-
a negative focus in relation to children;7 and (b) rate information on the frequency with which they
focusing on ‘emotional wellbeing’ in the broad spent time with their father who lived elsewhere.
sense is more consistent with prior work than an
idiosyncratic focus on ‘compromised’ emotional Analytic approach and background
wellbeing. For these reasons, our findings use the variables
language of children’s wellbeing being better or Bi-variate analyses were first conducted to explore
poorer in certain circumstances. (The distribu- the extent to which each of the three assessments of
tion of emotional difficulties scores is shown in children’s emotional wellbeing differed. Ordinary
Appendix Table 1. Correlations between the three Least Squares (OLS) regression was then applied
measures of children’s emotional wellbeing are to the data to assess the extent to which the three
reported in Appendix Table 2).8 sets of assessments differed according to the main
The main variables of interest in the following variables of interest when various background fac-
analyses are family type (whether child has a father tors were controlled.9
living elsewhere), the inter-parental hostility (hos- The background variables used in this analy-
tility vs no hostility), and father–child contact (no sis represent a selection of those that have already
face-to-face contact vs some contact). Five groups been shown to explain differences in the LSAC
of children are the focus of our analyses: children’s socio-emotional outcomes at age 4–5
and 6–7 years (Smart et al., 2008). The contri-
1. Children who were living with both parents, bution of these variables to children’s emotional
where parents had a hostile relationship with wellbeing was assessed using a stepwise approach,
each other; with child characteristics (age and sex) considered
2. Children who were living with both parents, first. This was followed by an assessment of the
where parents were not hostile with each other; additional contribution of the following infor-
3. Children who were living with one parent, mation provided by the mother:10 her assessment
where mothers reported a hostile relationship of financial circumstances (whether ‘just getting
with the father; along’ or ‘poor’, versus better financial wellbeing),

7
Psychometrically, scores had positively skewed distributions suggesting that the experience of poor emotional
wellbeing was fairly unusual.
8
All these correlations were fairly low, but significantly greater than zero. The highest correlation was derived for
scores provided by parents and teachers (r = 0.24), while the correlation between scores provided by children
and parents was somewhat higher than that between the scores provided by the children and teachers (r = 0.13
vs 0.09). These fairly low correlations suggest that there is considerable merit in exploring the views of these
three different informants. This divergence in reports is consistent with prior work (see, for example, Achenbach
et al., 1987; De Los Reyes & Kazdin, 2005; Loeber et al., 1991).
9
We treated the SDQ as a continuous measure in multivariate analyses for two reasons. First, other researchers
treat the SDQ as a continuous measure in multivariate analyses (see, for example, Edwards & Bromfield 2009;
Maynard & Harding, 2010). Second, we re-estimated the models using logistic regression (separating the
outcomes into those with high levels of emotional difficulties, and the rest) and found that the pattern of results
was consistent with the results using OLS. Goodman and Goodman (2009) recently noted that it is appropriate
to use the SDQ to identify gradations of child wellbeing, as opposed to identifying children with clinical levels of
difficulties.
10
As noted above, many fathers, including almost all who lived elsewhere, did not provide personal information.

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 91


Jennifer Baxter, Ruth Weston and Lixia Qu

TABLE 1: DISTRIBUTION OF CHILD, FAMILY AND MATERNAL CHARACTERISTICS BY FAMILY TYPE, QUALITY OF INTER-PARENTAL
RELATIONSHIP AND CONTACT WITH NON-RESIDENT FATHER

Lives with both parents Has father living elsewhere Total

Non-hostile Hostile inter- Non-hostile Hostile inter- No contact


inter-parental parental inter-parental parental with father
relationship relationship relationship relationship

Child characteristics
Age (months), mean 82 82 82 82 82 82
Boy (%) 51 53 47 57 52 51
Family characteristics
Just getting along or 18 25 44 52 59 24
poor
Mothers’ characteristics
Incomplete 17 16 28 22 31 18
secondary education
Secondary or post- 50 52 52 61 58 51
secondary, less than
bachelor degree
Bachelor degree or 33 32 20 17 11 31
higher
High distress 1 4 4 7 5 2
Low parental warmth 13 18 12 11 15 13
High angry parenting 9 19 11 17 22 12
Low inductive 19 23 20 12 16 19
reasoning with child
Low parental 18 30 21 31 30 21
consistency
Lives with step-father n.a. n.a. 24 6 8 2
Number of
observations
Sample size 2,816 763 356 227 75 4,303
Note: Sample sizes do not add to total as data on inter-parental relationship quality was not available for 66 respondents.
In this table the hostility of the inter-parental relationship between mother and father living elsewhere is derived for all
children who have some contact with their father.

her age (up to 30 years versus older), her edu- post-secondary qualifications other than a degree;
cation level (incomplete secondary education achievement of a bachelor degree or higher quali-
and no post-school qualifications; completion fication) and an indicator of the mother’s self-
of secondary education and/or achievement of reported level of distress.11

11
The six-item Kessler-6 screening scale for psychological distress was used. The K6 questions ask the respondent
to ask how often in the last four weeks they felt (a) ‘Nervous’; (b) ‘Hopeless’; (c) ‘Restless or fidgety’; (d) ‘So
depressed that nothing could cheer you up’; (e) ‘That everything was an effort’; and (f) ‘Worthless’. There were
five response categories, from ‘None of the time’ (‘0’) to ‘All the time’ (‘4’). These values were summed and those
with a sum of 13 or higher (out of a possible maximim of 24) were considerted to be of serious risk of mental
health problems. These respondents are identifed as ‘distressed’ in these analyses.

92 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

TABLE 2: EMOTIONAL WELLBEING BY FAMILY TYPE AND STATUS OF INFORMANT, CHILDREN AGED 6–7 YEARS, MULTIVARIATE
RESULTS (OLS)

Mother Teacher Child

(1) (2) (1) (2) (1) (2)

Coefficient and 95% confidence interval


Family type (Reference = lives
with both parents)
Father lives elsewhere 0.5*** 0.2* 0.5*** 0.3** 0.5*** 0.4***
[0.3, 0.7] [0.0, 0.4] [0.3, 0.7] [0.1, 0.5] [0.3, 0.7] [0.2, 0.6]
Factors controlled
+ Child variables Yes Yes Yes Yes Yes Yes
+ Parent characteristics and Yes Yes Yes
parenting style
*p<0.05, **p<0.01, ***p<0.001.
Scales for all informants range from 0 to 10, with higher scores reflecting poorer emotional wellbeing. Mothers’ and
teachers’ reports are based on the SDQ-emotional symptoms subscale, while the children’s self-reports are based on
ratings of how often they felt ‘scared or worried’, ‘sad’, ‘angry or mad’ and how often they felt ‘happy’ (reverse scored).
Full model results are available in Appendix.

Four measures of maternal parenting style were never’ to 5 ‘all the time’. Ratings for some items
also included, again based on self-report: ‘parental on the various parenting style measures were
warmth’, ‘angry parenting’, ‘inductive reasoning’, subsequently reversed, so that high ratings on all
and ‘consistent parenting’. The ‘Parental Warmth’ items within a scale consistently reflected the par-
measure focused on such issues as how often the enting style suggested by the scale’s name.12
respondents expressed affection, and how often The distribution of these scales was used to cre-
they had a warm and close time together with ate a dummy variable for each scale that discrimi-
their children. The ‘Angry Parenting’ scale gauged nated between mothers who indicated that their
such matters as how often they praised their chil- style of parenting was more negative (in relation
dren, how often they expressed disapproval of to ‘angry’ parenting) or less positive (in relation
their children’s behaviour, and how often they to warmth, reasoning and consistency) than most
felt angry when they punished their children. The other mothers. Specifically, mothers with scores in
‘Inductive Reasoning’ measure tapped, among the bottom 20% of the distributions for warmth,
other behaviour, the frequency with which moth- inductive reasoning and consistency were respec-
ers provided an explanation when they corrected tively treated as engaging in low parental warmth,
their children’s behaviour, and how often they inductive reasoning and consistency (taken sepa-
reasoned with their children when children mis- rately) relative to other mothers. Likewise, those
behaved. Finally, the ‘Consistency Parenting’ scale with scores in the top 20% for the Angry Parenting
assessed such issues as how often mothers made scale were treated as adopting a parenting style
sure that the children obeyed directions, and how that entailed relatively high levels of anger.
often their children ignored any punishment they We also examined the extent to which chil-
received. For each question, mothers rated their dren’s scores regarding emotional wellbeing were
behaviour on a scale ranging from 1 ‘never/almost related to their living with a step-father, net of the

12
For instance, mothers’ ratings of the frequency with which they praised their child were reversed scored, given
that this item formed part of the ‘angry parenting’ scale (see Baxter & Smart, 2010 for details on these measures).

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 93


Jennifer Baxter, Ruth Weston and Lixia Qu

Reported by mother Reported by teacher Reported by child


Mean emotional difficulties score 4.0

3.0
Intact
Father elsewhere

2.0

1.0

FIGURE 1: MEAN EMOTIONAL WELLBEING SCORES BY FAMILY TYPE AND STATUS OF INFORMANT, CHILDREN AGED 6–7 YEARS
Note: Scales for all informants range from 0 to 10, with higher scores reflecting poorer emotional wellbeing. Mothers’
and teachers’ reports are based on the SDQ-emotional symptoms subscale, while the children’s self-reports are based on
ratings of how often they felt ‘scared or worried’, ‘sad’, ‘angry or mad’ and how often they felt ‘happy’ (reverse scored).
Figure shows 95% Confidence Interval.

other background factors described above. This assessment. In addition, mothers with an appar-
variable was included because of the likely com- ently hostile inter-parental relationship were more
plexity of children’s experiences leading up to, likely than mothers in non-hostile relationships
and including, life in a stepfamily,13 and the fairly to report difficult financial circumstances (intact
consistent research findings that children raised in families: 25% vs 18%; separated families entailing
step-families tend to fare less well in terms of sev- at least monthly contact: 52% vs 44%).
eral aspects of wellbeing than those who are raised Further, compared with mothers in intact fami-
in intact families. (For a review of this literature, lies with a non-hostile inter-parental relationship,
including alternative explanations for such find- mothers in a hostile inter-parental relationship
ings, see Sweeney, 2010; Wu, Hou, & Schimmele, were more likely to report high distress, low paren-
2008.) Much of the relevant literature, however, tal warmth, high angry parenting, low parental
focuses on wellbeing by the time the children had consistency and low inductive reasoning. Three of
reached adolescence or young adulthood. these trends were also apparent among separated
mothers in a hostile inter-parental relationship,
RESULTS compared with mothers in a non-hostile relation-
ship: separated mothers in a hostile relationship
Background characteristics of the
were more likely than other separated mothers
five groups of children
to indicate that they displayed a parenting style
Table 1 shows the extent to which the background entailing anger and low consistency as well as a
factors (including parenting style) varied for five high level of distress. Nevertheless, it should be
key groups of children. Separated mothers were noted that reports of negative parental styles and
more likely than those who lived with their child’s distress were uncommon across all groups assessed.
father to see themselves as either ‘just getting along’,
‘poor’ or ‘very poor’ (reported by 44–59% of these Family type
respondents), with mothers whose child never saw Figure 1 shows the mean scores for chil-
the father being the most likely to provide this dren’s emotional wellbeing according to the

13
In many cases, children in step-families would have been spent time adjusting to life in a single parent families
before the mother re-partnered. Such experiences, along with possibly improved financial circumstances, and
the quality of their relationships with all three parent-figures (mother, father and stepfather) may well affect their
emotional wellbeing.

94 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

FIGURE 2: MEAN EMOTIONAL WELLBEING SCORES BY CONFLICT BETWEEN RESIDENT AND NON-RESIDENT PARENT
AND STATUS OF INFORMANT, CHILDREN AGED 6–7 YEARS WITH PARENT LIVING ELSEWHERE

Note: Scales for all informants range from 0 to 10, with higher scores reflecting poorer emotional wellbeing. Mothers’
and teachers’ reports are based on the SDQ-emotional symptoms subscale, while the children’s self-reports are based on
ratings of how often they felt ‘scared or worried’, ‘sad’, ‘angry or mad’ and how often they felt ‘happy’ (reverse scored).
Figure shows 95% Confidence Interval.

relationship status of the informant and family background factors were controlled [see column
type. (Background characteristics have not been of results with heading ‘(2)’].
taken into account here.) According to all infor- Both before and after these controls were
mants, children with a father living elsewhere had introduced, the assessments provided by moth-
poorer emotional wellbeing, compared with chil- ers and children suggest that, among children in
dren who lived with both biological parents. intact families, those whose parents had a hostile
Table 2 shows that when the selection of back- inter-parental relationship tended to have poorer
ground factors are controlled, the emotional well- emotional wellbeing than those whose parents did
being of children of separated parents continued not have a hostile relationship. No such variation
to be lower (i.e. children’s Emotional Difficulties emerged when scores concerning emotional well-
scores are higher) than those whose parents had being were based on teachers’ assessments.
not separated. This was apparent in the reports of Before the controls were introduced
all the informants. At this broad level, then, the (Figure 2), the results based on teachers’ reports
results suggest that young children of separated suggest that the children with separated par-
parents fared worse in terms of emotional wellbe- ents were more likely to display poorer emo-
ing than those who lived with both parents. tional wellbeing than other children, regardless
of whether or not the parents’ held a hostile
Hostility in inter-parental relationship relationship with each other. However, the dif-
Figure 2 shows the average scores (derived from ference between the teachers’ assessments of
the three informant sources) of children repre- children in intact families and of children who
sented in these five key analytic groups before the never saw their father was not statistically sig-
effects of the selected background factors were nificant.14 This pattern of results also emerged
controlled. Table 3 shows the extent to which when the selected child-related characteristics
these results held when the selected child-related were controlled, but the difference for intact ver-
factors were controlled [see column of results sus separated parents was no longer significant
with heading ‘(1)’], and then when all selected when all background factors were controlled.

14
This non-significant result could be explained by the fact that teachers’ assessments of children who never saw
their father varied greatly (as reflected in the long vertical bar).

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 95


Jennifer Baxter, Ruth Weston and Lixia Qu

TABLE 3: EMOTIONAL WELLBEING BY CONFLICT BETWEEN RESIDENT AND NON-RESIDENT PARENT AND STATUS OF
INFORMANT, CHILDREN AGED 6–7 YEARS, MULTIVARIATE RESULTS (OLS)

Family type and hostility Mother Teacher Child


of relationship between
parents

(1) (2) (1) (2) (1) (2)

Coefficient (relative to reference category) and 95% confidence interval


Reference = lives with
both parents, non-hostile
relationship
Lives with both parents, 0.5*** 0.3*** 0.0 −0.1 0.3** 0.2*
hostile relationship [0.4, 0.7] [0.2, 0.5] [−0.2, 0.1] [−0.3, 0.1] [0.1, 0.4] [0.0, 0.4]
Has father living elsewhere-
Non-hostile relationship 0.4** 0.2 0.5** 0.3 0.3 0.2
[0.1, 0.7] [−0.1, 0.5] [0.1, 0.8] [−0.0, 0.6] [−0.0, 0.6] [−0.1, 0.5]
Hostile relationship 0.7*** 0.3* 0.4** 0.3 0.9*** 0.8***
[0.4, 0.9] [0.0, 0.6] [0.1, 0.7] [−0.0, 0.6] [0.6, 1.2] [0.4, 1.1]
No contact 1.0** 0.6* 0.4 0.3 0.7** 0.4
[0.4, 1.6] [0.1, 1.2] [−0.2, 1.1] [−0.3, 1.0] [0.2, 1.2] [−0.1, 0.9]
Factors controlled
+ Child variables Yes Yes Yes Yes Yes Yes
+ Parental characteristics Yes Yes Yes
and parenting style
*p<0.05, **p<0.01, ***p<0.001.
Scales for all informants range from 0 to 10, with higher scores reflecting poorer emotional wellbeing. Mothers’ and
teachers’ reports are based on the SDQ-emotional symptoms subscale, while the children’s self-reports are based on
ratings of how often they felt ‘scared or worried’, ‘sad’, ‘angry or mad’ and how often they felt ‘happy’ (reverse scored).
Full model results are available in Appendix.

In other words, the differences in the teachers’ Much the same pattern of results emerged
assessments of children in intact and separated when all the selected controls were introduced,
families could be explained in terms of system- although trends based on all three informant
atic differences between these groups in the fam- groups suggested that children of separated par-
ily background factors examined. ents who did not have a hostile relationship did
Again, before the various controls were intro- not exhibit significantly poorer emotional wellbe-
duced, the assessments of all three informants sug- ing than their counterparts in intact families (i.e.,
gested that, for most comparisons, the children whose parents did not have a hostile relationship).
who lived with both parents who had a non-hos- In addition, according to children’s reports, the
tile relationship had better emotional wellbeing emotional wellbeing of children who never saw
than the three groups of children whose fathers their father was not significantly different to that
were living elsewhere. The only two exceptions of children who lived in an intact family entailing
to this finding related to teachers’ assessments for a non-hostile inter-parental relationship. A statis-
children who never saw their father and to the tically significant difference between these groups
reports of children whose parents had separated was apparent according to mothers’ reports, with
but held a non-hostile relationship. those having no face-to-face contact with their

96 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

father having poorer emotional wellbeing than the children whose parents had a hostile rela-
those who had contact. tionship, those whose parents had separated had
Before the controls were introduced, the only poorer wellbeing than those whose parents were
significant difference between the mean scores of living together, according to teachers’ reports,
the three groups of children whose father lived (p = 0.004 before controls and p = 0.02 after con-
elsewhere emerged in the children’s reports: of the trols) and children’s reports (p < 0.001 before con-
children who had some contact with their father, trols, p = 0.001 after controls), but not according
those whose parents had a hostile relationship to mothers’ reports.
with each other appeared to experience poorer
emotional wellbeing than those whose parents Hostility and frequency of contact
had a non-hostile relationship (p < 0.001). This Here we expand on the above analyses to incor-
difference remained significant when the con- porate, for children of separated parents, the fre-
trols were introduced (p = 0.004). As evident quency with which they spent time with their
in Figure 3, child emotional wellbeing appears father who lived elsewhere. Those who saw their
somewhat poorer for those children who never father at least weekly and those who saw him at
see their father, according to mothers’ reports, least monthly (including at least fortnightly) were
however, the mean emotional difficulties score for sub-divided according to whether or not the rela-
this group did not differ significantly from that of tionship between the parents was hostile. Those
other children with a father living elsewhere. In who saw their father less frequently were included
fact, regardless of the informant (mother, teacher in the analyses but not classified according to the
or child), the mean scores for children who never inter-parental hostility measure. This approach
saw their father did not differ significantly from yields six groups of children with fathers living
the mean scores for the other two groups of chil- elsewhere. Figure 3 provides the bi-variate results
dren whose father lived elsewhere. This was appar- regarding the six groups of children whose father
ent before and after the controls were introduced. lived elsewhere. Table 4, which summarises the
The analyses also allow other comparisons multivariate results, examines these six groups,
regarding the extent to which children in dif- but also includes the two groups of children
ferent circumstances examined were judged to whose parents had not separated: those whose
have different levels of emotional wellbeing. Of parents’ relationship was hostile or non-hostile.

FIGURE 3: MEAN EMOTIONAL WELLBEING SCORES BY PARENTAL CONFLICT AND FREQUENCY OF CONTACT,
AND STATUS OF INFORMANT, CHILDREN AGED 6–7 YEARS WITH FATHER LIVING ELSEWHERE

Note: Scales for all informants range from 0 to 10, with higher scores reflecting poorer emotional wellbeing. Mothers’
and teachers’ reports are based on the SDQ-emotional symptoms subscale, while the children’s self-reports are based on
ratings of how often they felt ‘scared or worried’, ‘sad’, ‘angry or mad’ and how often they felt ‘happy’ (reverse scored).
Figure shows 95% Confidence Interval.

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 97


Jennifer Baxter, Ruth Weston and Lixia Qu

When based on parents’ and teachers’ assess- experience of different parenting arrangements
ments of children’s emotional wellbeing, none deserve careful attention – especially given that
of the differences (depicted in Figure 3) in aver- most Australian research of patterns of parenting
age scores of the six groups of children (with after separation is based on parents’ reports, not
separated parents) was statistically significant. children’s reports.
However, one significant difference emerged According to the bi-variate analyses of moth-
in the children’s assessments: of those who saw ers’ and teachers’ assessments, no significant dif-
their father at least weekly, children whose par- ferences emerged in the emotional wellbeing
ents had a hostile relationship reported poorer scores of the six groups of children with separated
emotional wellbeing than those whose parents parents. However, in the multivariate analyses
did not have a hostile relationship. This differ- (summarised in Table 4), which included chil-
ence remained significant in the multivariate dren in intact families, significant differences
analyses, even when the controls were introduced emerged for some groups. For example, based
(Table 4, p < 0.002). Children’s reports of their on mothers’ reports: (a) among children in intact

TABLE 4: EMOTIONAL WELLBEING BY PARENTAL CONFLICT AND FREQUENCY OF CONTACT, AND STATUS OF INFORMANT,
CHILDREN AGED 6–7 YEARS, MULTIVARIATE RESULTS (OLS)

Frequency of contact with PLE Mother Teacher Child

(1) (2) (1) (2) (1) (2)

Coefficient (relative to reference category) and 95% Confidence Interval


(Lives with both parents, not
hostile)
Up to once a week, not hostile 0.6** 0.3 0.5** 0.4 0.1 0.0
[0.2, 1.0] [−0.1, 0.7] [0.1, 1.0] [−0.0, 0.8] [−0.3, 0.5] [−0.4, 0.4]
Up to once a week, hostile 0.6** 0.2 0.4* 0.3 1.1*** 0.9***
[0.2, 1.0] [−0.2, 0.6] [0.0, 0.9] [−0.1, 0.7] [0.6, 1.5] [0.5, 1.3]
Up to once a month, not hostile 0.2 0.1 0.4 0.3 0.5* 0.4
[−0.1, 0.6] [−0.2, 0.5] [−0.1, 0.9] [−0.2, 0.8] [0.0, 0.9] [−0.0, 0.8]
Up to once a month, hostile 0.7** 0.5* 0.4 0.3 0.7** 0.6*
[0.2, 1.2] [0.0, 0.9] [−0.1, 0.9] [−0.2, 0.7] [0.3, 1.2] [0.1, 1.0]
Less often 0.7** 0.3 0.5* 0.3 0.8*** 0.5*
[0.3, 1.1] [−0.1, 0.7] [0.0, 0.9] [−0.1, 0.8] [0.4, 1.2] [0.1, 1.0]
Never 0.9** 0.5 0.5 0.3 0.7** 0.5*
[0.3, 1.4] [−0.0, 1.0] [−0.1, 1.0] [−0.3, 0.9] [0.2, 1.1] [0.0, 0.9]
Lives with both parents, hostile 0.5*** 0.3*** 0.0 −0.1 0.3** 0.2*
[0.4, 0.7] [0.2, 0.5] [−0.2, 0.1] [−0.3, 0.1] [0.1, 0.4] [0.0, 0.4]
Factors controlled
+ Child variables Yes Yes Yes Yes Yes Yes
+ Parental characteristics & Yes Yes Yes
parenting style
*p<0.05, **p<0.01, ***p<0.001.
Scales for all informants range from 0 to 10, with higher scores reflecting poorer emotional wellbeing. Mothers’ and
teachers’ reports are based on the SDQ-emotional symptoms subscale, while the children’s self-reports are based on
ratings of how often they felt ‘scared or worried’, ‘sad’, ‘angry or mad’ and how often they felt ‘happy’ (reverse scored).
Full model results are available in Appendix.

98 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

families, those whose parents had a hostile rela- negatively. However, children of somewhat dis-
tionship tended to have poorer emotional well- tressed mothers did not appear to have signifi-
being than those whose parents did not have a cantly poorer emotional wellbeing according to
hostile relationship, and (b) the latter group of teachers’ assessments or according to the chil-
children seemed to fare better emotionally than dren themselves. This is consistent with find-
one group of children whose separated parents ings reported by Richters (1992) and suggests
held a hostile relationship: those who saw their that mothers who experience relatively high dis-
father at least monthly. As with the other analyses, tress either over-report the emotional difficulties
teachers’ reports yielded fewer differences in emo- experienced by their children or are better able
tional wellbeing than children’s or parents’ reports than other people, including their young child,
– none of which was significant after the inclusion to detect such problems in their child. This raises
of control variables. the question of whether mothers are likely to
have a better handle on their children’s emotional
Other parental and family wellbeing than the children themselves, and the
characteristics corollary of this question, whether children aged
Before concluding, we briefly summarise the 6–7 years are able to interpret their emotions
results for the control variables (presented in accurately and are willing to disclose them. It is
detail in Appendix Tables 3–5) because some of important to note in this regard that the children’s
these help explain the less favourable results that self-report measure was quite crude. Had a more
emerged for children of separated parents. Many sophisticated self-report measure been used with
of the other factors in these tables were stronger children, a greater consistency between children’s
predictors of children’s emotional wellbeing than and mothers’ reports may have emerged. Not sur-
inter-parental hostility and frequency of father– prisingly, separated mothers were likely to indi-
child contact. cate higher levels of distress than non-separated
We found that, according to parents’ and mothers (see Table 1). Such systematic differences
teachers’ assessments, the children of mothers may have contributed to the more negative assess-
whose saw themselves in difficult financial cir- ments of the child’s emotional wellbeing provided
cumstances (i.e., ‘just getting along’, ‘poor’ or by separated than non-separated mothers.15
‘very poor’) experienced poorer emotional well- Children’s wellbeing was also significantly
being. Interestingly, this trend was not evident associated with two aspects of mothers’ parent-
in children’s reports. Of course, it is unclear to ing style. Firstly, mothers who reported that
what extent young children understand fam- they engaged in relatively high ‘angry parenting’
ily finances. Separated mothers (single or re- tended to provide less favourable assessments
partnered) were more likely to indicate such of their child’s emotional wellbeing, compared
financial difficulties – a trend that would have with other mothers. However, this effect was not
contributed to, but did not fully explain, poorer apparent when the children’s emotional wellbe-
emotional wellbeing of children in such sepa- ing was based on reports of either the children
rated families. themselves or their teacher. Secondly, the children
Compared with other mothers, those who of mothers who indicated relatively high incon-
indicated relatively high emotional distress also sistency in their parenting tended to fare worse
assessed their child’s emotional wellbeing more in terms of emotional wellbeing, when the latter

15
While it also seems reasonable to suggest that the negative emotional impact of parental separation on children
may heighten mothers’ distress, it has already been shown that a significant relationship between maternal
distress and the child’s emotional wellbeing only emerged when the latter measure was based on mothers’
assessments.

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 99


Jennifer Baxter, Ruth Weston and Lixia Qu

measure was based on assessments of mothers and associations with emotional wellbeing that we have
children, but not when it was based on teachers’ observed may be due to the unobserved differences
assessments. in families that explain the parenting and family
In addition, children of mothers with a degree structure measures as well as the child outcomes.
or higher qualification appeared to have better While some studies have been able to take account
emotional wellbeing when measured by children’s of ‘unobserved heterogeneity’ in exploring links
and teacher’s reports but not mothers’ reports. between family structure and children wellbeing,
Further, boys had poorer emotional wellbeing the results from these studies are quite mixed and
than girls, although this difference was very slight some find an association persists after the unob-
and may not translate to meaningful differences served heterogeneity is taken into account (see
in everyday life. This effect was not apparent Björklund, Ginther, & Sundström, 2007).
when the emotional wellbeing measure was based Mothers’ parenting style was explored to shed
on parents’ or teachers’ assessments. light on the typical quality of the time they spend
There was no evidence that children’s emo- with their child. For reasons noted above, fathers’
tional wellbeing differed for those living with a parenting style was not examined, nor were more
step-father, compared with those living with a direct measures made of the quality of time that
single mother or with both biological parents. parents spend with their child. Analyses of associ-
Nevertheless, as already noted, previous research ations between fathering and children’s outcomes
has suggested that, compared with children in using the LSAC data by Baxter and Smart (2010)
intact families, those in step-families are at a showed that fathers’ parenting styles and the qual-
higher risk of poor psycho-social adjustment and ity of the co-parental relationship helped explain
poor academic performance, among other prob- variation in children’s social emotional outcomes,
lems (see Sweeney, 2010; Wu et al., 2008). This after taking account of the variation explained by
issue is worthy of ongoing investigation as the various child and family characteristics. This sug-
child grows older and as, potentially, more chil- gests that it will be worthwhile exploring such fac-
dren enter step-parent families. tors further in future analyses.
For separated parents, the measure of inter-
LIMITATIONS AND FUTURE OPPORTUNITIES parental hostility we used is quite crude and
As with any analyses into the relative importance based entirely on mothers’ assessments. Fathers
of different factors in explaining children’s emo- may hold quite different views. Furthermore, it
tional wellbeing, there are a number of limitations was not possible to identify the extent to which
in this work. One important limitation is that this parents attempted to shield the child from expo-
study could not explore the extent to which par- sure to inter-parental hostility.
enting arrangements or hostility caused emotional A key limitation of this set of analyses is that
difficulties in children. Our analysis is based on we were unable to report fathers’ evaluations of
cross-sectional data and correlation does not mean their child’s wellbeing, since only one parent (the
causation. Differences in parenting arrangements, primary care-giver) provided such assessments in
for example, might affect child wellbeing but the Waves 1 and 2.16 Whether or not this matters is
converse is also possible: children’s emotional dif- unclear. Previous research has shown that mothers’
ficulties might influence parental relationships. and fathers’ reports on children’s wellbeing tend to
Further, while our analysis included a number be somewhat discordant (e.g., Moreno, Silverman,
of family characteristics, we have not taken account Saavedra, & Phares, 2008), and we could expect
of all aspects of family functioning, and so the that this discordance may be even higher when

16
Given that mothers were usually the primary caregivers, we omitted cases where fathers had assumed this role.

100 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

the parents do not live together, since they tend an increasing awareness of the importance of
to observe their child in different settings. In later developmental considerations in research on chil-
waves of the study (in Wave 417 and subsequently) dren (see, for example, Magnuson and Berger
it will be possible to compare mothers’ and fathers’ 2009; Cashmore and Parkinson 2011; McIntosh,
reports on child wellbeing in intact families, and Smyth, Kelahar, Wells, & Long, 2010). We used
also to compare the reports of parents who live LSAC data to explore the links between young
elsewhere. Ex-couple dyadic perspectives are likely children’s emotional wellbeing, the quality of
to provide a richer understanding of children’s the co-parental relationship, and post-separation
emotional wellbeing, especially when children’s paternal involvement. We found that while chil-
self-reports are also taken into account. dren aged 6–7 years living with both parents gen-
The preceding analyses can be developed fur- erally had better emotional wellbeing than similar
ther in a number of ways. The logical next step is aged children living with one parent, inter-paren-
to extend the work to take advantage of the lon- tal hostility was significantly associated with (and
gitudinal study design by incorporating informa- an important factor explaining) young children’s
tion about these 6–7 year old children in Waves 1 emotional wellbeing. Specifically, regardless of
and 3 – that is when the children were 4–5 years family type, children whose parents had a hostile
old and 8–9 years old – and taking into account inter-parental relationship tended to have poorer
changes that occurred in the inter-parental rela- emotional wellbeing than children whose parents
tionship and family type. Making use of the lon- did not have a hostile relationship, as reported by
gitudinal data would enable us to determine the children and their parents.
extent to which children exhibited emotional Amato (2000) found considerable overlap
problems before their parents separated, or even in the wellbeing of children in intact and sepa-
before the parents’ relationship became some- rated families. This is consistent with the findings
what hostile. Further, longitudinal analysis would of this paper, suggesting that most children are
enable us to begin to examine to the nature and quite resilient, and appear to cope well when con-
strength of longer term effects of parental sepa- fronted with major changes to their lives (Amato,
ration. Previous research (Cherlin et al., 1998) 2001; Amato & Rezac, 1994; Furstenberg &
has shown that separation often has many ripple Kiernan, 2001; Hetherington, 1999; Kelly &
effects, with some effects not becoming apparent Emery, 2003). And, of course, some children in
until adulthood (e.g., increased risk of early part- intact families exhibit quite marked psycho-emo-
nering and relationship breakdown). tional difficulties.
Finally, these analyses can – and should – be One interesting finding concerned children
repeated for different dimensions of children’s from separated families who saw their father
developmental progress, including cognitive on at least a weekly basis. Differences in emo-
development, physical health and other aspects of tional wellbeing, reported by these children,
socio-emotional development. were significantly linked to the level of inter-
parental hostility. This suggests (perhaps not
DISCUSSION AND CONCLUSION surprisingly) that more frequent time spent with
It is well known that children who experience fathers whose former partners report hostile rela-
parental divorce are more likely than those in tionships exposes the child to this hostility more
intact families to experience a range of emotional often. At the same time, however, children who
and behavioural adjustment problems. Yet few never saw their father also had poorer emotional
studies have focussed on young children, despite wellbeing than those in intact families in which

17
In Wave 4, conducted in 2010, the children were 10–11 years old.

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 101


Jennifer Baxter, Ruth Weston and Lixia Qu

the parents did not have a hostile relationship child to detect signs of internalised emotional
with each other. This is consistent with recent problems.
Australian analyses of children’s wellbeing in It needs to also be acknowledged that measur-
separated families (Kaspiew et al., 2009; Qu & ing psychological domains is a complex pursuit.
Weston, 2010). It suggests that we need to know Indeed, variables predicting mothers’ and chil-
more about the ‘absent father’ group. In some dren’s assessments were not always consistent:
situations, such as cases when the father engages for example financial difficulties, and moth-
in systemic violence or other dysfunctional ers’ heightened level of distress (both of which
behaviours, children’s interests may be bet- were reported by mothers) were associated with
ter served by no father–child contact until and children displaying poorer emotional wellbeing
unless the father is willing to change his behav- according to mothers, but not according to chil-
iour. But father absence may nonetheless leave dren. Having a mother with a degree or higher
mothers more vulnerable. And regardless of the qualification was associated with better emotional
reasons, father absence may be perceived by the wellbeing according to children’s assessments, but
child as abandonment. not according to mothers’ assessments. On the
There was clear evidence that, on average, other hand, wellbeing appeared to be more com-
children with separated parents have poorer promised from both the perspective of mothers
emotional wellbeing than those in intact fami- and children where the mother engaged in a more
lies. Some of this difference is likely to reflect inconsistent parenting style, and as noted above,
systematic differences in other characteristics where the parents had a hostile relationship with
of separated and intact families – for example, each other.
separated families were more likely to experience It is possible that different views of infor-
difficult financial circumstances and a higher mants are all picking up some valid differences
proportion of mothers reported the adoption in aspects of emotional wellbeing of children in
of negative parenting styles. But associations different settings. The different findings based on
between variables such as these and children’s the assessments of mothers, teachers and children
wellbeing apply also in intact families, such that have implications for identifying children with
these variables have a significant association with emotional difficulties. Importantly, to the extent
children’s wellbeing even when the different that the children’s self-report measure concern-
measures of co-parenting and post-separation ing emotional experiences has an acceptable level
paternal involvement are included in the analy- of validity, it would seem that some children are
ses. For instance, mothers’ self-reported incon- likely to miss out on the emotional help they
sistent approaches to parenting were a significant need, where such needs are based exclusively on
predictor of negative emotional wellbeing in the assessments of mothers or teachers. The mea-
children, where the latter was reported by the sure based on children’s assessments was crude
child or assessed by the mothers. and the validity of this measure needs further test-
An important contribution of this paper is ing. Clearly however, children’s own assessments
the use of data from three different informants. have the potential to provide important insights
Of the three assessments of children’s emotional into their internal worlds that may otherwise go
wellbeing, those provided by teachers seemed to unnoticed.
have the weakest associations with the charac- Finally, we note the emerging evidence
teristics explored here. This suggests that chil- (Cashmore & Parkinson, 2011) that a sizeable
dren’s emotional wellbeing may be less apparent proportion of recently separated families involve
in the classroom. In addition, perhaps teachers young children (i.e., children under 5). Some of
are not sufficiently familiar with each individual these separations will serve to relieve the conflict

102 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

between the parents. But many will not – and Amato, P. R. (2000). The consequences of divorce for
indeed some will entail an intensification of con- adults and children. Journal of Marriage and the
Family, 62, 1269–1287.
flict. Traditionally, parents often stayed together Amato, P. R. (2001). Children of divorce in the 1990s:
for the sake of the children. But the devastating An update of the Amato and Keith (1991)
impact of conflict on children was not well rec- meta-analysis. Journal of Family Psychology, 15(3),
ognised. In Australia, a raft of new family rela- 355–370.
Amato, P. R. (2010). Research on divorce: Continuing
tionship services was introduced in 2006 and trends and new developments. Journal of Marriage
some existing services were expanded. A key aim and Family, 72(3), 650–666.
of many of these services has been to assist both Amato, P. R., & Booth, A. (1997). A generation at risk:
‘intact’ and separated families to manage conflict Growing up in an era of family upheaval. Cambridge,
MA: Harvard University Press.
within families, including that between the par-
Amato, P. R., & Gilbreth, J. G. (1999).
ents. From the point of view of some children, Nonresident fathers and children’s well-being: A
separation may often be seen as an undesirable meta-analysis. Journal of Marriage and the Family, 61,
outcome of inter-parental conflict. At the same 557–573.
Amato, P. R., & Rezac, S. J. (1994). Contact with
time, our analysis of LSAC data suggests that
nonresident parents, interparental conflict, and
efforts by family relationship services such as children’s behavior. Journal of Family Issues, 15(2),
Family Relationship Centres to assist parents to 191–207.
reduce or overcome their hostility towards each Ambert, A. M. (1997). Parents, children, and adolescents:
other are likely to benefit young children irrespec- Interactive relationships and development in context.
London: Routledge.
tive of whether those parents remain together or Australian Bureau of Statistics. (2001). Yearbook Australia
separate. (Catalogue No. 1301.0). Canberra, Australian Capital
Territory: Author.
ACKNOWLEDGEMENTS Australian Bureau of Statistics. (2007). 2006 Census of
We are grateful to the reviewers for their construc- population and housing: Census table (Catalogue No.
2068.0). Canberra, Australian Capital Territory:
tive comments. An earlier version of this paper Author.
was presented at the XXVI International Union Australian Bureau of Statistics. (2011). Family character-
for the Scientific Study of Population (IUSSP) istics Australia 2009–2010 (Catalogue No. 4442.0).
International Population Conference, Marrakech Canberra, Australian Capital Territory: Author.
Australian Institute of Family Studies. (2005).
Morocco. This paper uses unit record data from
Growing up in Australia the longitudinal study of
Growing Up in Australia, the Longitudinal Study Australian children annual report. Melbourne,
of Australian Children. The study is conducted in Victoria: Author.
partnership between the Department of Families, Baxter, J. A., & Smart, D. (2010). Fathering in Australia
among couple families with young children (FaHCSIA
Housing, Community Services and Indigenous
Occasional Paper No. 37). Canberra, Australian
Affairs (FaHCSIA), the Australian Institute of Capital Territory, Australia: FaHCSIA.
Family Studies (AIFS) and the Australian Bureau Björklund, A., Ginther, D. K., & Sundström, M. (2007).
of Statistics (ABS). The findings and views Family structure and child outcomes in the USA
reported in this paper are those of the authors and and Sweden. Journal of Population Economics, 20(1),
183–201.
should not be attributed to FaHCSIA, AIFS or Briggs-Gowan, M., Carter, A., & Schwab-Stone, M.
the ABS. (1996). Discrepancies among mother, child, and
teacher reports: Examining the contributions of
References maternal depression and anxiety. Journal of Abnormal
Achenbach, T. M., McConaughy, S. H., & Howell, C. T. Child Psychology, 24(6), 749–765.
(1987). Child/adolescent behavioral and emotional Cashmore, J., & Parkinson, P. (2011) Parenting
problems: Implications of cross-informant correla- arrangements for young children: Messages
tions for situational specificity. Psychological Bulletin, from research. Australian Journal of Family Law, 25,
101(2), 213–232. 1–22.

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 103


Jennifer Baxter, Ruth Weston and Lixia Qu

Carlson, M. J., & Corcoran, M. E. (2001). Family child mental health. Journal of the American Academy
structure and children’s behavioral and cognitive of Child & Adolescent Psychiatry, 48(4), 400–403.
outcomes. Journal of Marriage and the Family, 63(3), Goodman, R. (2001). Psychometric properties of the
779–792. Strengths and Difficulties Questionnaire. Journal
Cherlin, A. J. (2008). Multiple partnerships and of the American Academy of Child & Adolescent
children’s wellbeing. Family Matters, 80, 33–36. Psychiatry, 40(11), 1337–1345.
Cherlin, A. J., Chase-Lansdale, P. L., & McRae, C. Gray, M., & Smart, D. (2008). Growing up in
(1998). Effects of parental divorce on mental health Australia: The longitudinal study of Australian
throughout the life course. American Sociological children is now walking and talking. Family Matters,
Review, 63, 239–249. 79, 5–13.
Chi, T., & Hinshaw, S. (2002). Motherñchild relation- Hayes, A., Weston, R., Qu, L., & Gray, M. (2010).
ships of children with ADHD: The role of mater- Families then and now, 1980–2010 (Facts Sheet).
nal depressive symptoms and depression-related Melbourne, Victoria: Australian Institute of Family
distortions. Journal of Abnormal Child Psychology, Studies.
30(4), 387–400. Hetherington, E. (1999). Should we stay together for the
Choudhury, M. S., Pimentel, S. S., & Kendall, P. C. sake of the children. In E. Hetherington (Ed.), Coping
(2003). Childhood anxiety disorders: Parent-child with divorce, single parenting, and remarriage: A risk
(dis) agreement using a structured interview for the and resiliency perspective (pp. 93–116). Mahwah, NJ:
DSM-IV. Journal of American Academy of Child & Lawrence Erlbaum.
Adolescent Psychiatry, 42(8), 957–964. Jekielek, S. M. (1997). Parental conflict, marital disrup-
Cole, D., Hoffman, K., Tram, J., & Maxwell, S. (2000). tion and children’s emotional well-being. Social Forces,
Structural differences in parent and child reports 76(3), 905–936.
of children’s symptoms of depression and anxiety. Karver, M. S. (2006). Determinants of multiple
Psychological Assessment, 12(2), 174–185. informant agreement on child and adolescent
Comer, J., & Kendall, P. (2004). A symptom-level exami- behavior. Journal of Abnormal Child Psychology, 34(2),
nation of parent-child agreement in the diagnosis of 242–253.
anxious youths. Journal of American Academy of Child Kaspiew, R., Gray, M., Weston, R., Moloney, L., Hand,
& Adolescent Psychiatry, 43(7), 878–886. K., & Qu, L. (2009). Evaluation of the 2006 family
De Los Reyes, A., & Kazdin, A. (2005). Informant law reforms. Melbourne, Victoria: Australian Institute
discrepancies in the assessment of childhood psycho- of Family Studies.
pathology: A critical review, theoretical framework, Kelly, J. B., & Emery, R. E. (2003). Children’s adjust-
and recommendations for further study. Psychological ment following divorce: Risk and resilience perspec-
Bulletin, 131(4), 483–509. tives. Family Relations, 52(4), 352–362.
Duhig, A., Renk, K., Epstein, M., & Phares, V. (2000). Lodge, J., & Alexander, M. (2010). Views of adolescents
Interparental agreement on internalizing, external- in separated families: A study of adolescents’ experi-
izing, and total behavior problems: A meta-analysis. ences after the 2006 reforms to the family law system.
Clinical Psychology: Science and Practice, 7(4), Melbourne, Victoria: Australian Institute of Family
435–453. Studies.
Edwards, B., & Bromfield, L. M. (2009). Neighborhood Loeber, R., Green, S. M., Lahey, B. B., & Stouthamer-
influences on young children’s conduct problems Loeber, M. (1991). Differences and similarities
and pro-social behavior: Evidence from an Australian between children, mothers, and teachers as infor-
national sample. Children and Youth Services Review, mants on disruptive child behavior. Journal of
31(3), 317–324. Abnormal Child Psychology, 19(1), 75–95.
Emery, R. E. (1982). Interparental conflict and the Magnuson, K., & Berger, L. M. (2009). Family structure
children of discord and divorce. Psychological Bulletin, states and transitions: Associations with children’s
92(2), 310–330. well-being during middle childhood. Journal of
Furstenberg Jr, F. F., Morgan, S. P., & Allison, P. D. Marriage and the Family, 71(3), 575–591.
(1987). Paternal participation and children’s well- Maynard, M. J., & Harding, S. (2010). Ethnic differ-
being after marital dissolution. American Sociological ences in psychological well-being in adolescence
Review, 52(5), 695–701. in the context of time spent in family activities.
Furstenberg, F. F., Jr., & Kiernan, K. (2001). Delayed Social Psychiatry and Psychiatric Epidemiology, 45(1),
parental divorce: How much do children benefit? 115–123.
Journal of Marriage and the Family, 63(2), 446–457. McIntosh, J., Smyth, B., Kelahar, M., Wells, Y., & Long,
Goodman, A., & Goodman, R. (2009). Strengths and C. (2010) Post-separation parenting arrangements
difficulties questionnaire as a dimensional measure of and developmental outcomes for infants and children:

104 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

Collected reports. Melbourne, Victoria, Australia: Sarrazin, J., & Cyr, F. (2007). Parental conflicts and their
Family Transitions. damaging effects on children. Journal of Divorce &
Measelle, J. R., Ablow, J. C., Cowan, P. A., & Cowan, Remarriage, 47(1), 77–93.
C. P. (1998). Assessing young children’s views Silverman, W., & Rabian, B. (1995). Test-retest reli-
of their academic, social, and emotional lives: ability of the DSM-III-R childhood anxiety disorders
An evaluation of the self-perception scales of the symptoms using the anxiety disorders interview
Berkeley Puppet Interview. Child Development, 69(6), schedule for children. Journal of Anxiety Disorders,
1556–1576. 9(2), 139–150.
Moreno, J., Silverman, W. K., Saavedra, L. M., & Smart, D., Sanson, A., Baxter, J., Edwards, B., & Hayes,
Phares, V. (2008). Fathers’ ratings in the assessment A. (2008). Home-to-school transitions for financially
of their child’s anxiety symptoms: A comparison to disadvantaged children. Sydney, New South Wales:
mothers’ ratings and their associations with paternal The Smith Family.
symptomatology. Journal of Family Psychology, 22(6), Smyth, B. (2004). Parent–child contact schedules after
915–919. divorce. Family Matters 69, 32–43.
Morrison, D. R., & Coiro, M. J. (1999). Parental Soloff, C., Lawrence, D., & Johnstone, R. (2005). Sample
conflict and marital disruption: Do children benefit design (LSAC Technical Paper No. 1). Melbourne,
when high-conflict marriages are dissolved? Journal of Victoria: Australian Institute of Family Studies.
Marriage and the Family, 61, 626–637. Sweeney, M. M. (2010). Remarriage and stepfamilies:
Pryor, J., & Rodgers, B. (2001). Children in changing Strategic sites for family scholarship in the 21st
families: Life after parental separation: Cambridge, century. Journal of Marriage and Family, 72, 667–684.
MA: Blackwell. Whiteside, M., & Becker, B. (2000). Parental factors
Qu, L., & Weston, R. (2010). Parenting dynamics after and the young child’s postdivorce adjustment: A
separation: A follow-up study of parents who sepa- meta-analysis with implications for parenting arrange-
rated after the 2006 family law reforms. Melbourne, ments: Marriage and divorce: Correlates, conse-
Victoria: Australian Institute of Family Studies. quences, and processes. Journal of Family Psychology,
Richters, J. (1992). Depressed mothers as informants 14(1), 5–26.
about their children: A critical review of the evidence Wu, Z., Hou, F., & Schimmele, C. M. (2008). Family
for distortion. Psychological Bulletin, 112, 485–485. structure and children’s psychosocial outcomes.
Riley, A. (2004). Evidence that school-age children can Journal of Family Issues, 29(12), 1600–1624.
self-report on their health. Ambulatory Pediatrics,
4(4), 371–376. Received 21 June 2011 Accepted 02 August 2011

APPENDIX
TABLE 1: FREQUENCY DISTRIBUTION OF SAMPLE BY INTER-PARENTAL HOSTILITY AND FREQUENCY OF CONTACT OF CHILD
WITH PLE

Distribution – sample count Not hostile Hostile Total

Lives with both parents 2,816 763 3,640


Lives with mother; father lives elsewhere: Spends 286 247 663
time with father:
Weekly or more frequently 134 112 258
At least fortnightly or monthly 87 80 184
Less frequently 54 48 121
Not at all 11 7 100
Total 3,102 1,010 4,303
Columns do not add to total as some respondents did not provide sufficient information to determine whether hostile or not.

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 105


Jennifer Baxter, Ruth Weston and Lixia Qu

TABLE 2: CHILD’S SCORES ON THE COMPROMISED EMOTIONAL WELLBEING MEASURES BY INFORMANT STATUS: SUMMARY
STATISTICS, INTERNAL CONSISTENCY, AND INTER-CORRELATIONS

Mean (SD) Median Cronbach’s alpha Correlations between measures

Mother Teacher Child

Mother – SDQ emotional 1.6 (1.7) 1 0.62 1.00


subscale (5 items)
Teacher – SDQ emotional 1.3 (1.8) 1 0.74 0.24 1.00
subscale (5 items)
Child – self-report emotional 3.1 (1.9) 2.5 0.42 0.13 0.09 1.00
scale (3 items)

TABLE 3: MULTIVARIATE RESULTS FOR FAMILY TYPE, OLS COEFFICIENTS AND 95% CONFIDENCE INTERVALS
Mother Teacher Child

(1) (2) (1) (2) (1) (2)

(Reference = Lives with both parents)


Father lives elsewhere 0.5*** 0.2* 0.5*** 0.3** 0.5*** 0.4***
[0.3, 0.7] [0.0, 0.4] [0.3, 0.7] [0.1, 0.5] [0.3, 0.7] [0.2, 0.6]
Child lives with −0.2 −0.1 −0.1 −0.1 −0.1 −0.1
step-father [−0.7, 0.2] [−0.4, 0.3] [−0.7, 0.4] [−0.7, 0.4] [−0.5, 0.4] [−0.5, 0.4]
Child’s characteristics
Age (months divided by 0.2 0.1 0.0 −0.1 −0.1 −0.1
10, centred at sample [−0.0, 0.3] [−0.1, 0.3] [−0.2, 0.2] [−0.3, 0.2] [−0.3, 0.1] [−0.3, 0.1]
mean)
Boy −0.1 −0.1 −0.1 −0.1 0.1* 0.2**
[−0.2, 0.1] [−0.2, 0.0] [−0.2, 0.0] [−0.3, 0.0] [0.0, 0.3] [0.0, 0.3]
Mother’s report re financial status
Just getting along, poor 0.3*** 0.3*** 0.0
or very poor [0.2, 0.5] [0.1, 0.5] [−0.1, 0.2]
Mother’s characteristics:
Age ≤ 30 years 0.2* 0.0 0.2
[0.0, 0.4] [−0.3, 0.2] [−0.0, 0.4]
Completed or post- 0.0 −0.1 −0.2*
secondary, less than [−0.2, 0.1] [−0.2, 0.1] [−0.3, 0.0]
degreea
Bachelor degree or −0.1 −0.2* −0.5***
higher [−0.3, 0.1] [−0.4, −0.0] [−0.7, −0.3]
Parenting style: low 0.2 0.0 0.1
warmth [−0., 0.3] [−0.3, 0.2] [−0.1, 0.4]
Parenting style: high 0.7*** 0.1 0.1
angry [0.5, 0.9] [−0.1, 0.4] [−0.1, 0.3]
Parenting style: low −0.2 −0.1 −0.1
inductive reasoning [−0.3, 0.0] [−0.3, 0.0] [−0.3, 0.0]

(Continued)

106 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

TABLE 3: CONTINUED
Mother Teacher Child

(1) (2) (1) (2) (1) (2)

Parenting style: low 0.3*** 0.0 0.3***


consistency [0.2, 0.5] [−0.2, 0.1] [0.1, 0.4]
High distress 1.2*** 0.4 0.4
[0.7, 1.7] [−0.1, 0.9] [−0.0, 0.8]
Constant 1.6*** 1.4*** 1.3*** 1.3*** 3.0*** 3.0***
[1.5, 1.7] [1.2, 1.6] [1.2, 1.4] [1.1, 1.4] [3.0, 3.1] [2.8, 3.1]
N 4,186 4,159 3,484 3,410 4,216 4,094
R-square 0.01 0.08 0.01 0.02 0.01 0.03
*p<0.05, **p<0.01, ***p<0.001.
Child’s age is a continuous variable, centred at the mean of the sample. Maternal education is relative to incomplete
secondary education. ‘Completed or post-secondary, less than degree’ refers to the mother either having completed
secondary education or having achieved a post-school education other than a degree. Other parental and family
characteristics are dummy variables, relative to not having that characteristic. Column ‘(1)’ refers to results when child’s
characteristics were entered; Column ‘(2)’ refers to results when all listed characteristics were entered.

TABLE 4: MULTIVARIATE RESULTS FOR FAMILY TYPE AND INTER-PARENTAL HOSTILITY, OLS COEFFICIENTS AND 95%
CONFIDENCE INTERVALS

Parent Teacher Child

(1) (2) (1) (2) (1) (2)

(Reference category = Lives with both parents, non-hostile inter-parental relationship)


Lives with both parents, 0.5*** 0.3*** 0.0 −0.1 0.3** 0.2*
hostile inter-parental relp [0.4, 0.7] [0.2, 0.5] [−0.2, 0.1] [−0.3, 0.1] [0.1, 0.4] [0.0, 0.4]
Father elsewhere, non- 0.4** 0.2 0.5** 0.3 0.3 0.2
hostile inter-parental relp [0.1, 0.7] [−0.1, 0.5] [0.1, 0.8] [−0.0, 0.6] [−0.0, 0.6] [−0.1, 0.5]
Father elsewhere, hostile 0.7*** 0.3* 0.4** 0.3 0.9*** 0.8***
inter-parental relp [0.4, 0.9] [0.0, 0.6] [0.1, 0.7] [−0.0, 0.6] [0.6, 1.2] [0.4, 1.1]
Father elsewhere, no 1.0** 0.6* 0.4 0.3 0.7** 0.4
contact [0.4, 1.6] [0.1, 1.2] [−0.2, 1.1] [−0.3, 1.0] [0.2, 1.2] [−0.1, 0.9]
Lives with a step-father −0.2 0.0 0.2 0.2 0.0 0.0
[−0.7, 0.3] [−0.5, 0.4] [−0.4, 0.8] [−0.4, 0.8] [−0.5, 0.5] [−0.6, 0.5]
Child’s characteristics
Age (months divided by 10, 0.2 0.1 0.0 −0.1 −0.1 −0.1
centred at sample mean) [−0.0, 0.3] [−0.1, 0.3] [−0.2, 0.2] [−0.3, 0.2] [−0.3, 0.1] [−0.3, 0.1]
Boy 0.0 −0.1 −0.1 −0.1 0.1* 0.1*
[−0.2, 0.1] [−0.2, 0.0] [−0.2, 0.1] [−0.3, 0.0] [0.0, 0.3] [0.0, 0.3]
Mother’s report re financial status
Just getting along, poor or 0.3*** 0.3*** 0.0
very poor [0.2, 0.5] [0.1, 0.5] [−0.1, 0.2]
Mother’s characteristics:
Age ≤ 30 years 0.2* 0.0 0.2
[0.0, 0.4] [−0.3, 0.2] [−0.0, 0.4]

(Continued)
Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 107
Jennifer Baxter, Ruth Weston and Lixia Qu

TABLE 4: CONTINUED
Parent Teacher Child

(1) (2) (1) (2) (1) (2)

Completed or post- 0.0 −0.0 −0.2


secondary, less than degree [−0.2, 0.2] [−0.2, 0.1] [−0.3, 0.0]
Bachelor degree or higher −0.1 −0.2* −0.5***
[−0.2, 0.0] [−0.3, −0.0] [−0.7, −0.3]
Parenting style: low warmth 0.1 0.0 0.2
[−0.0, 0.3] [−0.2, 0.2] [−0.1, 0.4]
Parenting style: high angry 0.6*** 0.1 0.1
[0.4, 0.8] [−0.1, 0.4] [−0.1, 0.3]
Parenting style: low −0.2* −0.2 −0.1
inductive reasoning [−0.3, −0.0] [−0.3, 0.0] [−0.3, 0.1]
Parenting style: low 0.3*** 0.0 0.3***
consistency [0.1, 0.4] [−0.2, 0.1] [0.1, 0.4]
High distress 1.2*** 0.4 0.3
[0.6, 1.7] [−0.1, 1.0] [−0.2, 0.7]
Constant 1.5*** 1.3*** 1.3*** 1.3*** 3.0*** 3.1***
[1.4, 1.6] [1.2, 1.5] [1.2, 1.4] [1.1, 1.4] [2.9, 3.1] [2.9, 3.3]
N 4,115 4,088 3,406 3,358 4,102 4,027
R-square 0.03 0.08 0.01 0.02 0.02 0.03
*p<0.05, **p<0.01, ***p<0.001.
Child’s age is a continuous variable, centred at the mean of the sample. Maternal education is relative to incomplete
secondary education. ‘Completed or post-secondary, less than degree’ refers to the mother either having completed
secondary education or having achieved a post-school education other than a degree. Other parental and family
characteristics are dummy variables, relative to not having that characteristic. Column ‘(1)’ refers to results when child’s
characteristics were entered; Column ‘(2)’ refers to results when all listed characteristics were entered.

TABLE 5: MULTIVARIATE RESULTS FOR FAMILY TYPE, INTER-PARENTAL HOSTILITY AND FREQUENCY WITH WHICH CHILD SEES
NON-RESIDENT FATHER, OLS COEFFICIENTS AND 95% CONFIDENCE INTERVALS

Parent Teacher Child

(1) (2) (1) (2) (1) (2)

(Reference = Lives with both parents, non-hostile inter-parental relationship)


Lives with both parents, 0.5*** 0.3*** 0.0 −0.1 .3** .2*
hostile [0.4, 0.7] [0.2, 0.5] [−0.2, 0.1] [−0.3, 0.1] [0.1, 0.4] [0.0, 0.4]
Father lives elsewhere: frequency of time and inter-parental relationship
Up to once a week, 0.6** 0.3 0.5** 0.4 .1 .0
non-hostile [0.2, 1.0] [−0.1, 0.7] [0.1, 1.0] [−0.0, 0.8] [−0.3, 0.5] [−0.4, 0.4]
Up to once a week, hostile 0.6** 0.2 0.4* 0.3 1.1*** .9***
[0.2, 1.0] [−0.2, 0.6] [0.0, 0.9] [−0.1, 0.7] [0.6, 1.5] [0.5, 1.3]
Up to once a month, 0.2 0.1 0.4 0.3 .5* .4
non-hostile [−0.1, 0.6] [−0.2, 0.5] [−0.1, 0.9] [−0.2, 0.8] [0.0, 0.9] [−0.0, 0.8]
Up to once a month, 0.7** 0.5* 0.4 0.3 .7** .6*
hostile [0.2, 1.2] [0.0, 0.9] [−0.1, 0.9] [−0.2, 0.7] [0.3, 1.2] [0.1, 1.0]

(Continued)

108 JOURNAL OF FAMILY STUDIES Volume 17, Issue 2, August 2011


Children’s emotional wellbeing

TABLE 5: CONTINUED
Parent Teacher Child

(1) (2) (1) (2) (1) (2)

Less often 0.7** 0.3 0.5* 0.3 .8*** .5*


[0.3, 1.1] [−0.1, 0.7] [0.0, 0.9] [−0.1, 0.8] [0.4, 1.2] [0.1, 1.0]
Never 0.9** 0.5 0.5 0.3 .7** .5*
[0.3, 1.4] [−0.0, 1.0] [−0.1, 1.0] [−0.3, 0.9] [0.2, 1.1] [0.0, 0.9]
Lives with step-father −0.3 −0.1 −0.1 −0.1 .0 .0
[−0.7, 0.1] [−0.5, 0.3] [−0.7, 0.4] [−0.7, 0.4] [−0.5, 0.5] [−0.5, 0.5]
Child’s characteristics
Age (months divided by 10, 0.1 0.1 0.0 −0.1 −0.1 −0.1
centred at sample mean) [−0.0, 0.3] [−0.1, 0.3] [−0.3, 0.2] [−0.3, 0.2] [−0.4, 0.1] [−0.3, 0.1]
Boy −0.1 −0.1 −0.1 −0.1 .1* .1*
[−0.2, 0.1] [−0.2, 0.0] [−0.2, 0.1] [−0.3, 0.0] [0.0, 0.2] [0.0, 0.3]
Mother’s report re financial status:
Just getting along, poor or 0.3*** 0.3*** .1
very poor [0.2, 0.5] [0.1, 0.5] [−0.1, 0.2]
Mother’s characteristics
Age ≤ 30 years 0.2* 0.0 .2
[0.0, 0.4] [−0.3, 0.2] [−0.0, 0.4]
Complete or post- 0.0 0.1 .2*
secondary, less than degree [−0.2, 0.2] [−0.1, 0.2] [0.0, 0.3]
Bachelor degree or higher −0.1 −0.2* −0.3***
[−0.2, 0.0] [−0.3,−0.0] [−0.5,−0.2]
Parenting style: low warmth 0.1 0.0 .1
[−0.0, 0.3] [−0.2, 0.2] [−0.1, 0.4]
Parenting style: high angry 0.6*** 0.2 .1
[0.4, 0.9] [−0.1, 0.4] [−0.1, 0.3]
Parenting style: low −0.1* −0.2 −0.1
inductive reasoning [−0.3,−0.0] [−0.3, 0.0] [−0.3, 0.1]
Parenting style: low 0.3*** 0.0 .3***
consistency [0.1, 0.4] [−0.2, 0.1] [0.1, 0.4]
High distress 1.2*** 0.4 .3
[0.7, 1.7] [−0.1, 0.9] [−0.2, 0.7]
Constant 1.5*** 1.3*** 1.3*** 1.3*** 3.0*** 2.9***
[1.4, 1.6] [1.2, 1.4] [1.2, 1.4] [1.2, 1.4] [2.9, 3.1] [2.8, 3.1]
N 4152 4126 3433 3385 4137 4062
R-square 0.03 0.08 0.01 0.02 0.02 0.04

Child’s age is a continuous variable, centred at the mean of the sample. Maternal education is relative to incomplete
secondary education. ‘Completed or post-secondary, less than degree’ refers to the mother either having completed
secondary education or having achieved a post-school education other than a degree. Other parental and family
characteristics are dummy variables, relative to not having that characteristic. Column ‘(1)’ refers to results when child’s
characteristics were entered; Column ‘(2)’ refers to results when all listed characteristics were entered.

Volume 17, Issue 2, August 2011 JOURNAL OF FAMILY STUDIES 109


Copyright of Journal of Family Studies is the property of eContent Management Pty. Ltd. 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.

You might also like