Structural Equation Model to Study the Mediation Effect of Maternal Co-construction in Mother-child Attachment Relationship

• In Event: Poster Session 03
  In Poster Session: PS 03 Topic - Parenting & Parent-Child Relationships

Abstract

Attachment theory hypothesizes that children’s attachment knowledge is acquired from their interactions with parents and represented mentally as a model that organizes information obtained and aids children in interpreting and navigating interactions. It has been shown that those representations are not simply derived from behavioral exchanges, but actively constructed through verbal communications that help (or not) preschoolers organize their attachment-related knowledge (e.g., Lu et al., 2018). Those authors found that verbal co-construction skills contribute to the organization of child security beyond any contributions of maternal sensitivity during child-mother interactions. Yet, the point in development at which maternal attachment-related verbalizations begin to play a role on security outcomes is unknown. Are those demonstrated verbal contributions to child security well into the preschool-years characteristic of dyads with younger preschoolers (e.g., 2/3-year-olds)? This has important implications for theory and practice (e.g., intervention programs). We investigated whether maternal sensitivity and verbal co-construction skills predict child’s secure base use (security) at 2-3 years of age and whether those skills mediate the sensitivity-security association.
Fifty-three families with their 29.2-month-olds (27 girls) completed all assessments. Mother’s mean age was 31.8 years; 65% declared themselves as the main caregiver, 31% declared both parents as the main caregivers, and 4% declared father as the main caregiver. Mother-child interaction was observed at home and park (75 minutes each) by trained observers. Maternal sensitivity was assessed with the Maternal Behavior for Preschoolers Q-set (Posada et al., 2007), child’s secure base use (security) was assessed with the Attachment Q-set (Waters, 1995). Observers for maternal and child behavior were independent (inter-observer reliabilities were > .70 for both). Dyads came to the lab and completed a joint story telling task; their interaction was recorded and used to code maternal co-construction skills when discussing attachment-related events. Three 7-point scales (“Creating a co-construction atmosphere,” “Encouraging content elaboration,” and “Supporting an explanatory framework”) were used to assess maternal co-construction skills (Posada & Waters, 2018). Mothers were rated on those scales by trained coders (alpha= .92), independent from those who assessed behaviors at home and park.
Pearson correlations among the variables are presented in Table 1. We used structural equation modeling and created a latent variable, maternal co-construction skills, by loading the three co-construction scales. Confirmatory factor analysis for the measurement model indicated good fit, t=142.284, df=3, p=.00, CFI=TLI=1, RMSEA=.00. Path model analysis (Figure 1) using sensitivity as independent variable, co-construction as mediator and security as the dependent variable showed fair model fit, t=167.609, df=10, p=.00, CFI=.958, TLI=.895, RMSEA=.149. Sensitivity significantly predicted child security; however, co-construction skills did not predict security and the mediation effect was not significant.
Results indicate that maternal sensitivity remains an important factor as security outcomes are concerned early in the preschool years. However, maternal verbal communications were not found associated with sensitivity in the path model or child security. This suggests that verbal influences in security, found later in the preschool years, may not come into play until the child representational/verbal abilities are further developed and dyadic verbal exchanges more characteristic.

Authors

• Muqing Liu, Purdue University, West Lafayette
  Presenting Author

• Geurim Kim, Purdue University, West Lafayette
  Non-Presenting Author

• Megan Bieber, Western Illinois University
  Non-Presenting Author

• Samuel Morehead, Purdue University
  Non-Presenting Author

• German E Posada, Purdue University, West Lafayette
  Non-Presenting Author