Attachment from infancy to preschool: Domestic violence and other predictors of stability and instab

1
Attachment from infancy to preschoolDomestic
violence and other predictors of stability and
instability

• Alytia A. Levendosky and G. Anne Bogat
• Michigan State University
• Alissa C. Huth-Bocks
• Eastern Michigan University
• Katherine L. Rosenblum
• University of Michigan

2
Collaborators and Grant Support

• Collaborators
• Alexander von Eye, Ph.D. Mariam Mourad, M.A.
• William Davidson, Ph.D. Archana Basu, M.A.
• Sally Theran, Ph.D. Cecilia Martinez-Torteya,
  B.A.
• Robin Weatherill, Ph.D. Parker Huston, B.A.
• Erika DeJonghe, M.A. Lia Field, B.A.
• Kerry Leahy, M.A.
• Carolyn Dayton, M.A.
• Grant support
• NIJ 2 years of funding
• CDC 3 years of funding

3
Background

• Attachment
• Stability and instability
• Low-risk versus high-risk
• Predictors of attachment stability
• Domestic violence
• Income
• Maternal caregiving representations
• Gender

4
Hypotheses/Research Questions

• What is the concordance rate of attachment in
  this heterogeneous for risk group from ages 1 to
  4?
• How does the trajectory of DV influence the
  change/stability of the childs attachment
  classifications from ages 1 to 4?
• How does the trajectory of Income influence the
  change/stability of the childs attachment
  classifications from ages 1 to 4?

5
Hypotheses/Research Questions

• How does the change/stability of Maternal
  Representations from prenatal to 1 year postnatal
  affect the change/stability of attachment from
  ages 1 to 4?
• How is Gender related to attachment
  change/stability from ages 1 to 4?

6
Longitudinal Study

• Research design
• Quasi-experimental DV vs no DV
• Began during pregancy
• Followed yearly until children age 7
• Sample
• 206 pregnant women
• Initial recruitment
• Convenience sampling from community

7
Demographics

• Women
• 62 Caucasian, 25 African-American
• 47 single, 43 married
• 42 high school or less, 44 trade school or some
  college
• Children
• 72 males, 78 females
• 47 Caucasian, 25 African-American, 24
  Multiracial

8
Measures

• Demographics
• Income, Gender of child
• Severity of Violence Against Women
• Total severity score
• Working Model of the Child Interview
• Balanced, Nonbalanced
• Strange Situation
• Secure, Insecure

9
Procedures

• Interview
• Lab observations
• Home interviews
• Tracking
• Every 3 months contact
• Protocol for letters, calling, visiting
• Payment for contact response
• 20 attrition over 9 waves (7 ½ years)
• Coding
• WMCI
• Strange Situation

10
Attachment Classifications
11
Patterns of Attachment

• Secure-Secure n 58
• Secure-Insecure n 26
• Insecure-Secure n 38
• Insecure-Insecure n 28

12
Results Domestic Violence Trajectory

• Repeated Measures ANOVA
• 5 periods of DV (prenatal, age 1, age 2, age 3
  and age 4)
• Significant time by pattern interaction
• Linear effect of time all groups decreased in
  DV over time
• Quadratic effect of time by pattern such that
  there were different curvature for the
  trajectories
• Secure-insecure group had relatively low rates of
  DV with a concave trajectory
• Insecure-insecure group had highest rates of
  prenatal DV with a convex trajectory

13
Results Income Trajectory

• Repeated Measures ANOVA
• 5 time periods (prenatal, age 1, age 2, age 3,
  and age 4)
• Significant effect of time by pattern
• Cubic effect of time by pattern
• Secure-secure group had significantly higher
  income over all times compared with both
  insecure-secure and insecure-insecure groups
• Insecure-secure group began with relatively low
  income which steadily and slowly rose over time

14
Results Maternal Representations

• Using chi-sq and Configural Frequency Analysis
• Significant relationship between pattern of
  maternal representations and patterns of
  attachment
• Types
• secure-secure and balanced-balanced
• unstable attachment and unstable maternal
  representations
• unstable attachment and nonbalanced-nonbalanced
• Insecure-insecure and balanced-nonbalanced

15
Results - Gender

• Significant difference in attachment
  classifications by gender
• More girls secure at age 1 and more boys secure
  at age 4
• No interaction with DV
• Interaction with Maternal Representations
• Stability and instability of maternal
  representations predicted stability or
  instability of attachment for boys but not for
  girls

16
Discussion

• Attachment concordance not significant
• Predictors
• Domestic violence disrupts mother-child
  relationship and/or is associated with emotional
  dysregulation in the child
• Low income (i.e. economic distress) disrupts
  parenting behavior and affects mother-child
  relationship

17
Discussion

• Predictors (continued)
• Stability of maternal balanced representations
  was a protective factor for stability of secure
  attachment
• Gender was associated with change in maternal
  representations and subsequently in attachment
  classifications
• Experiences with male infant positively
  influenced maternal representations

18
Limitations and Conclusions

• All maternal report
• Domestic violence trajectory
• Support for broader theory about influence of
  environmental factors on stability and
  instability of attachment