Differential Susceptibility to Family Conflict: The Interaction Effect of Temperament

1
Differential Susceptibility to Family Conflict
The Interaction Effect of Temperament
Parenting on Early Childhood Delinquent Symptoms

• Xiao Zhang
• Department of Psychology
• University of Jyväskylä

2
Background

• Part of a longitudinal project of Chinese
  children during the transition to nursery care

3
Background

• Part of a longitudinal project of Chinese
  children during the transition to nursery care
• Transition to child care
• ---inherently stressful for young children
  (unfamiliar adults, unknown peers, and new
  physical settings for extended periods of time in
  absence of their parents)

4
Background

• Part of a longitudinal project of Chinese
  children during the transition to nursery care
• Transition to child care
• ---inherently stressful for young children
  (unfamiliar adults, unknown peers, and new
  physical settings for extended periods of time in
  absence of their parents)
• ---increased risk of
  psychopathology symptoms across cultures (Ahnert
  et al., 2004 Klein et al., 2010 Zhang Sun,
  2010)

5
Background

• Part of a longitudinal project of Chinese
  children during the transition to nursery care
• Transition to child care
• ---inherently stressful for young children
  (unfamiliar adults, unknown peers, and new
  physical settings for extended periods of time in
  absence of their parents)
• ---increased risk of
  psychopathology symptoms across cultures (Ahnert
  et al., 2004 Klein et al., 2010 Zhang Sun,
  2010)
• Initial intention Family School ?
  Psychopathology
• (Temperament
  covariate)

6
Background

• Part of a longitudinal project of Chinese
  children during the transition to nursery care
• Transition to child care
• ---inherently stressful for young children
  (unfamiliar adults, unknown peers, and new
  physical settings for extended periods of time in
  absence of their parents)
• ---increased risk of
  psychopathology symptoms across cultures (Ahnert
  et al., 2004 Klein et al., 2010 Zhang Sun,
  2010)
• Initial intention Family School ?
  Psychopathology
• (Temperament
  covariate)
• Aim of this paper Temperament Family ?
  Psychopathology

7
Literature Review
8
Family Conflict

• Characterized by marital discord and more general
  anger, hostility, and criticism among family
  members.

9
Family Conflict

• Characterized by marital discord and more general
  anger, hostility, and criticism among family
  members.
• A marker of dysfunctional families (Cummings,
  1994).
• ---associated with elevated risk for a wide
  range of psychopathology symptoms, including
  internalizing (e.g., withdrawal, depression,
  anxiety) and externalizing (e.g., aggression,
  delinquency) (see a review in Kelly, 2000)

10
Family Conflict

• Not all children who experience conflictual and
  harsh home environments develop symptoms of
  psychopathology (Garmezy et al., 1984).

11
Family Conflict

• Not all children who experience conflictual and
  harsh home environments develop symptoms of
  psychopathology (Garmezy et al., 1984).
• A number of child characteristics have been
  suggested to moderate (either exacerbate or
  attenuate) the risk of stressful family
  environments

12
Family Conflict

• Not all children who experience conflictual and
  harsh home environments develop symptoms of
  psychopathology (Garmezy et al., 1984).
• A number of child characteristics have been
  suggested to moderate (either exacerbate or
  attenuate) the risk of stressful family
  environments
• ---behavioral or temperamental in character
  (e.g., difficult/easy temperament), physiological
  or endophenotypic in nature (e.g., physiological
  reactivity), and genetic in origin (e.g.,
  5-HTTLPR alleles) (Belsky Pluess, 2009)

13
Temperament

• Biologically based individual differences in
  reactivity and self-regulation, influenced over
  time by heredity, maturation, and experience
  (Rothbart Bates, 1998).

14
Temperament

• Biologically based individual differences in
  reactivity and self-regulation, influenced over
  time by heredity, maturation, and experience
  (Rothbart Bates, 1998).
• Difficult temperament
• ---operationalized in many different ways
  (e.g., high activity, negative emotionality)
• ---repeatedly linked to the development of
  internalizing and externalizing symptoms during
  the preschool period (Fagan, 1990)

15
Temperament

• Biologically based individual differences in
  reactivity and self-regulation, influenced over
  time by heredity, maturation, and experience
  (Rothbart Bates, 1998).
• Difficult temperament
• ---operationalized in many different ways
  (e.g., high activity, negative emotionality)
• ---repeatedly linked to the development of
  internalizing and externalizing symptoms during
  the preschool period (Fagan, 1990)
• Yet recent research suggests that the combination
  of difficult temperament with stressful family
  environments (or the Temperament Environment
  interaction) often has the greatest impact on
  psychopathology symptoms (e.g., Ramos et al.,
  2005 Tchann et al., 1996 Whiteside-Mansell et
  al., 2009).

16
Temperament Family Conflict
17
Temperament Family Conflict

• Theoretical models (Temperament Environment)
• ---Diathesis-stress hypothesis (Monroe
  Simons, 1991)
• ---Differential susceptibility hypothesis
  (Belsky Pluess, 2009)

18
Temperament Family Conflict

• Theoretical models (Temperament Environment)
• ---Diathesis-stress hypothesis (Monroe
  Simons, 1991) children with difficult
  temperament are more vulnerable to stressful
  environments, showing more symptoms of
  psychopathology, than those with easy temperament
  
• ---Differential susceptibility hypothesis
  (Belsky Pluess, 2009)

19
Temperament Family Conflict

• Theoretical models (Temperament Environment)
• ---Diathesis-stress hypothesis (Monroe
  Simons, 1991) children with difficult
  temperament are more vulnerable to stressful
  environments, showing more symptoms of
  psychopathology, than those with easy temperament
  
• ---Differential susceptibility hypothesis
  (Belsky Pluess, 2009) difficult children who
  are most adversely affected by an environmental
  stressor are the very same ones who reap the most
  benefit from environmental support and
  enrichment, including the absence of adversity

20
Temperament Family Conflict

• Theoretical models (Temperament Environment)

21
Temperament Family Conflict

• Both hypotheses have received empirical support.

22
Temperament Family Conflict

• Both hypotheses have received empirical support.
• ---Diathesis-stress hypothesis In a
  cross-sectional study of 145 preschool children,
  children with difficult temperament (e.g., high
  activity, negative mood) in high-conflict
  families had the most internalizing and
  externalizing symptoms, whereas children with
  easy temperament had the fewest symptoms
  regardless of levels of family conflict (Tschann
  et al., 1996)

23
Temperament Family Conflict

• Both hypotheses have received empirical support.
• ---Diathesis-stress hypothesis In a
  cross-sectional study of 145 preschool children,
  children with difficult temperament (e.g., high
  activity, negative mood) in high-conflict
  families had the most internalizing and
  externalizing symptoms, whereas children with
  easy temperament had the fewest symptoms
  regardless of levels of family conflict (Tschann
  et al., 1996)
• ---Differential susceptibility hypothesis
  In a longitudinal study of 728 children from ages
  1 to 8, children with difficult temperament
  (e.g., inadaptability, unpredictability) were
  more likely to develop externalizing symptoms as
  a result of high conflict in the family but also
  tended to show fewer symptoms in low-conflict
  families than those with easy temperament
  (Whiteside-Mansell et al., 2009)

24
Temperament Family Conflict

• No support for either hypothesis

25
Temperament Family Conflict

• No support for either hypothesis In a
  longitudinal study of 337 children from ages 5 to
  17, no interaction of temperament and marital
  conflict was found in the prediction of
  internalizing or externalizing problems (Leve et
  al., 2005).

26
Temperament Family Conflict

• Possible Sources of the mixed results

27
Temperament Family Conflict

• Possible Sources of the mixed results
• ---reporter bias and shared method
  invariance the same informant to measure both
  temperament and symptoms

28
Temperament Family Conflict

• Possible Sources of the mixed results
• ---reporter bias and shared method
  invariance the same informant to measure both
  temperament and symptoms
• ---hard to compare the results across
  studies difficult temperament was almost always
  conceptualized in general senses (instead of
  using specific dimensions) but operationalized in
  different ways (i.e., using different measures)

29
Temperament Family Conflict

• Possible Sources of the mixed results
• ---reporter bias and shared method
  invariance the same informant to measure both
  temperament and symptoms
• ---hard to compare the results across
  studies difficult temperament was almost always
  conceptualized in general senses (instead of
  using specific dimensions) but operationalized in
  different ways (i.e., using different measures)
• ---direction of causality some studies
  assessed conflict, temperament, and symptoms
  concurrently

30
Temperament Family Conflict

• Thus, more studies are needed with (a)
  measurement of conflict, temperament, and
  behavior using different informants, (b)
  assessment of specific temperamental dimensions,
  and (c) controls for prior symptoms when
  examining the Temperament Family Conflict
  interaction in the prediction of later symptoms.

31
Temperament Family Conflict

• Moreover, the extent to which findings derived
  from studies of Western children apply to Chinese
  children during a critical transitional period is
  not clear.

32
Temperament Family Conflict

• Moreover, the extent to which findings derived
  from studies of Western children apply to Chinese
  children during a critical transitional period is
  not clear.
• ---Chinese childrens symptoms of
  psychopathology are associated with both
  temperament (e.g., shyness-sensitivity Chen et
  al., 1992, 2009a behavioral inhibition Chen et
  al., 2009b effortful control, positive
  emotionality, anger-irritability Zhou et al.,
  2009) and family environments (Phillips et al.,
  1991). Yet we do not know whether family
  environments interact with temperament to
  influence symptoms among Chinese children.

33
Temperament Family Conflict

• Moreover, the extent to which findings derived
  from studies of Western children apply to Chinese
  children during a critical transitional period is
  not clear.
• ---Chinese childrens symptoms of
  psychopathology are associated with both
  temperament (e.g., shyness-sensitivity Chen et
  al., 1992, 2009a behavioral inhibition Chen et
  al., 2009b effortful control, positive
  emotionality, anger-irritability Zhou et al.,
  2009) and family environments (Phillips et al.,
  1991). Yet we do not know whether family
  environments interact with temperament to
  influence symptoms among Chinese children.
• ---Symptoms of psychopathology during the
  transition to child care are affected by both
  environmental factors such as parenting (Klein et
  al., 2010) and child temperament (e.g.,
  approach/withdrawal, adaptability Ahnert et al.,
  2004). However, we do not know whether symptoms
  during this transition can be predicted by the
  Temperament Family Conflict interactions.

34
The Present Study

• The present study extends previous work by
  examining the Temperament Family Conflict
  interaction in the prediction of delinquent
  symptoms among Chinese children during a critical
  developmental periodthe transition from home to
  child care.

35
The Present Study

• The present study extends previous work by
  examining the Temperament Family Conflict
  interaction in the prediction of delinquent
  symptoms among Chinese children during a critical
  developmental periodthe transition from home to
  child care.
• To this end, we followed up a group of children
  in Beijing during their one-year transition to
  nursery care.

36
The Present Study

• Hypothesis a significant Temperament Family
  Conflict interaction

37
The Present Study

• Hypothesis a significant Temperament Family
  Conflict interaction
• ---this interaction is expected to be
  accounted for by the differential susceptibility
  hypothesis. Specifically, children with difficult
  temperament were expected to be more susceptible
  to the negative consequences of high conflict in
  the family, showing more delinquent symptoms, and
  also more positively affected by low conflict,
  showing fewer symptoms, than children with easy
  temperament.

38
Method
39
Method

• Participants
• ---At 3 months after nursery entry (Time 1
  ?T1?), 118 children (56 boys, 62 girls M 33.3
  months old, SD 3.1 months) participated with
  mothers, fathers, and head caregivers.
• ---At 9 months after T1 (Time 2 ?T2? i.e.,
  the end of nursery care), 11 children had
  transferred to other nurseries and did not
  participate. Attrition analyses did not reveal
  significant differences in the childs age, sex,
  delinquent symptoms, temperament, or family
  conflict at T1.

40
Method

• Measures

41
Method

• Measures
• ---Delinquent Symptoms at T1 and T2 via
  caregiver report using the delinquency subscale
  in Achenbachs (1992) Child Behavior
  Checklist/23 (CBCL/23). Alpha coefficients were
  0.75 and 0.79 for delinquent problems at T1 and
  T2, respectively. Higher scores indicate more
  symptoms.

42
Method

• Measures
• ---Delinquent Symptoms at T1 and T2 via
  caregiver report using the delinquency subscale
  in Achenbachs (1992) Child Behavior
  Checklist/23 (CBCL/23). The scale measures the
  extent to which children do not do what social
  rules require. Alpha coefficients were 0.75 and
  0.79 for delinquent problems at T1 and T2,
  respectively. Higher scores indicate more
  symptoms.
• ---Family Conflict At T1 via father report
  using the conflict subscale in Moos and Mooss
  (1986) Family Environment Scale (FES). The scale
  measures the amount of openly expressed anger and
  conflict among family members. The alpha
  coefficient was 0.67. Higher scores indicate a
  conflictual environment in the family.

43
Method

• Measure
• ---Temperament at T1 via mother reports
  using the activity level, rhythmicity, and
  approach/withdrawal subscales in Thomas and
  Chesss (1977) Child Temperament Questionnaire
  (CTQ). Approach/withdrawal reflects a childs
  tendency to approach new people and situations.
  Activity level measures a child's general level
  of energy and movement. Rhythmicity reflects the
  degree of predictability of the timing of the
  childs biological functions such as hunger,
  sleep-wake cycles, and elimination. Alpha
  coefficients were 0.69 for the rhythmicity scale,
  0.67 for the approach-withdrawal scale, and 0.63
  for the activity level scale. Lower scores in
  rhythmicity and approach/withdrawal and higher
  scores in activity level indicate difficult
  temperament.

44
Results

• Hierarchical regression to predict T2 delinquent
  symptoms (covariates T1 delinquent symptoms,
  maternal education, childs age and sex)

Predictors B (S.E.) ß ?R2
Family conflict .38 (.20) .17 .02
Activity level .06 (.24) .02
Rhythmicity .37 (.22) .16 .05
Approach/withdrawal .60 (.23) .26
Activity level Conflict .46 (.22) .21
Rhythmicity Conflict .36 (.26) .14 .05
Approach/withdrawal Conflict .28 (.21) .13
45
Results
B.51, p lt .01b
B.40, p lt .05b
46
Discussion

• Consistent with the diathesis-stress hypothesis,
  the results showed that children with difficult
  temperament (i.e., high activity level) were more
  vulnerable to family conflict and showed more
  delinquent symptoms as compared to children with
  relatively easy temperament (i.e., low activity
  level).

47
Discussion

• Consistent with the diathesis-stress hypothesis,
  the results showed that children with difficult
  temperament (i.e., high activity level) were more
  vulnerable to family conflict and showed more
  delinquent symptoms as compared to children with
  relatively easy temperament (i.e., low activity
  level).
• Moreover, as an important additional finding,
  children with difficult temperament were also
  influenced more positively by low conflict in
  their family than children with easy temperament.
  Difficult children showed fewer delinquent
  symptoms when reared in less conflictual
  families.

48
Discussion

• Consistent with the diathesis-stress hypothesis,
  the results showed that children with difficult
  temperament (i.e., high activity level) were more
  vulnerable to family conflict and showed more
  delinquent symptoms as compared to children with
  relatively easy temperament (i.e., low activity
  level).
• Moreover, as an important additional finding,
  children with difficult temperament were also
  influenced more positively by low conflict in
  their family than children with easy temperament.
  Difficult children showed fewer delinquent
  symptoms when reared in less conflictual
  families.
• In sum, these findings were consistent with the
  differential susceptibility hypothesis (Belsky
  Pluess, 2009).

49
Discussion

• Temperament Family Conflict interactions
  accounted for 5 of the variance in delinquent
  symptoms.
• This effect size is generally consistent with
  results reported in other studies investigating
  Temperament Parenting interactions in the
  development of externalizing problems. Moreover,
  the interaction effects were independent of prior
  delinquent symptoms, maternal education, and the
  childs age and sex.

50
Discussion

• Strengths
• ---different informants significant
  interactions cannot be solely ascribed to shared
  method variances.
• ---a longitudinal design was adopted
  inferences about direction of effects can be
  made.
• ---the Temperament Environment interaction
  was first tested in a sample of Chinese children
  during the transition to nursery care.

51
Discussion

• Limitations
• ---sample characteristics attrition rates
  were moderate, and families from higher SES
  backgrounds were overrepresented, which may
  restrict the generalizability of the study.
• ---measurement of temperament only three
  temperamental dimensions were measured in a
  single context with maternal reports.

52
Discussion

• Moderation by gender?
• Temperament-Environment correlation?

53
Conclusion

• This study provides empirical evidence for the
  childrens differential susceptibility to family
  environments hypothesis using a sample of Chinese
  children during their transition to nursery care.
  
• More specifically, our results confirmed the
  hypothesis that children with difficult
  temperament are more susceptible to family
  conflict, for better and for worse Compared with
  children with easy temperament, they showed more
  delinquent symptoms when exposed to family
  conflict, whereas they showed fewer symptoms in
  the absence of conflict, independent of maternal
  education and the childs age and sex.

54
Thank you!Wish you a pleasant time in Finland!