Does attachment style influence social support or the other way around? A longitudinal study of Early Head Start mothers

Abstract

Understanding the association between attachment style and social support is important for informing programs that seek to improve outcomes for families by intervening with either or both of these systems. The present study examines whether increasing levels of social support among 181 low-income, primarily African American mothers leads to changes in their self-reported attachment style, or whether attachment style influences the extent to which they perceive others as supportive. Results suggest that whereas scores on the avoidant attachment dimension were relatively stable and led to decreasing perceptions of social support over time, scores on the anxious dimension were more malleable, at least under conditions of low stress. For mothers who experienced fewer stressful life events, increasing social support led to decreased attachment anxiety over time. However, when life stress was high, social support had no such positive influence. Implications for the need to attend to mothers' attachment styles in providing appropriate and effective intervention are discussed.

Keywords:

• attachment
• social support
• parenting
• high risk families

Acknowledgements

This manuscript is dedicated to the late Carol McAllister, Co-Principal Investigator of the Pittsburgh Early Head Start Research project, long-time friend and colleague. The findings reported here are based on research conducted as part of the national Early Head Start Research and Evaluation Project funded by the Administration for Children and Families (ACF), US Department of Health and Human Services, through Grant #90-YF-0025 to the University of Pittsburgh. The research is part of the independent research the University of Pittsburgh conducted with the Family Foundations Early Head Start program, one of 17 programs participating in the national Early Head Start study. The authors are members of the Early Head Start Research Consortium, which consists of representatives from 17 programs participating in the evaluation, 15 local research teams, the evaluation contractors, and ACF. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government.

Notes

1. Thirty-five percent were missing data at any one time point; 15% were missing any two time points, and 3% had data at all three time points but were missing variable-level data.

2. Treatment effects were not expected at the 14-month assessment since many mothers had only received a few months of service (average length of service at 14 months was about five months). It was also not entirely surprising that treatment effects were not seen for the program group at 36 months, for several reasons. First, there were considerable difficulties in implementing this EHS program during the first two years of the study, as the site struggled to adapt to Head Start program guidelines. The EHS National Study (Love at al., 2002), of which this site was a part, found that program effects for the national study were significantly more positive for those sites that were fully implemented during the first year of the study, a standard not achieved by this site. Further, concurrent ethnographic work at this site suggests that there were numerous instances of “spillover” of treatment from the program group to comparison group participants that occurred because of the program's strong emphasis on community development and involvement (McAllister, Green, Terry, Herman, & Mulvey, 2003).