Poster #96 - Latent Transitions in Preschoolers’ Internal Working Models of Attachment Security

• In Event: Poster Session 14

Abstract

Internal working models (IWM) are the general representations that reflect children’s blueprint of the social world around them (Fivush, 2006). From an attachment perspective, early relationships inform security (or insecurity) in children’s IWM based on the degree to which caregivers and others can be expected to provide safety, comfort, and support in times of distress (Bretherton, 1985). Early adversity (e.g., poverty, maltreatment) is considered a threat to IWM security because adversity tends to correspond with disruptions in early relationships (Doyle & Cicchetti, 2017).
Limited work has evaluated children’s IWM using longitudinal designs. Further, it remains unclear how or whether children’s IWM change during the preschool years in contexts of adversity. To address this gap, we examined latent transitions over one year in preschool-aged children’s attachment IWM in a sample of maltreating and demographically similar, low-income nonmaltreating families. Further, we evaluated effects of maltreatment and a brief parenting intervention (supporting maltreating mothers’ abilities to reminisce in an elaborative and emotionally supportive manner) on IWM transitions.
Participants were 242 3- to 6-year-old children (n=83 nonmaltreated, n=80 maltreated control, n=79 maltreated intervention). At Time 1 (T1) and one year later (T2), children completed the MacArthur Story Stem Battery (MSSB; Bretherton et al., 1990) to assess their IWM of the mother-child relationship. MSSB videos were coded for five IWM elements: avoidance, behavioral dysregulation, maternal incompetence, attachment insecurity, and conflict escalation (kw/ICCs>.68).
Using the five indicators, initial latent class analyses were conducted at each time point in Mplus Version 8. Model fits of 2-, 3-, and 4-class solutions were compared and revealed that a 3-class solution fit the T1 and T2 data best. Thus, a random intercept latent transition analysis (Muthén & Asparouhov, 2022) was conducted using a 3-class solution (BIC=7539.09, T1 and T2 entropy=.86). The three resulting classes reflected secure (nT1=132, nT2=205), insecure avoidant (nT1=77, nT2=20), and insecure dysregulated (nT1=33, nT2=17) IWM.
Children in the secure class at T1 had a 93.2% chance of remaining in the secure class at T2, showing high stability in secure IWM. Distinctly, children in the insecure avoidant and insecure dysregulated classes had a 29.6% and 13.4% chance of remaining in their initial class at T2, respectively, revealing relative instability in insecure IWM. Latent transition probabilities revealed that children in the insecure avoidant class at T1 had a statistically significant likelihood of transitioning to the secure class at T2 (probability=57.7%, odds=1.95, 95% CI: 1.047, 3.626) and children in the insecure dysregulated class at T1 had a statistically significant likelihood of transitioning to the secure class at T2 (probability=81.3%, odds=6.05, 95% CI: 1.812, 20.215). Covariate effects revealed that maltreatment and intervention group status did not significantly predict latent class transitions.
Overall, over a one-year period, children tended to maintain or transition towards security in their IWM. These results suggest that even in contexts of severe adversity children show propensities to form secure IWM. Future work should evaluate developmental mechanisms and protective factors that may explain why some children do or do not develop secure IWM over time.

Authors

• Ruth Speidel, University of Toronto
  Presenting Author

• Monica Lawson, University of Texas at San Antonio

• Brigid M Behrens, University of Notre Dame

• Katherine Edler, University of Notre Dame

• Kristin Valentino, University of Notre Dame