Date of Degree


Document Type


Degree Name





Deborah L. Vietze

Subject Categories

Developmental Psychology | Psychology


attachment; childhood; internal working model; maternal sensitivity; SECCYD; social behavior


Chapter 1 attempts to clarify the internal working model (IWM) by offering a revised and more precise definition of the concept. Chapter 2 discusses the existing IWM research and asserts what aims should be applied to future research to further clarify the IWM concept. Chapters 3 and 4 describe the current three studies that sought to test the validity of the proposed IWM definition by determining if: (1) attachment status during the first three years fluctuates as a function of changes in maternal sensitivity, (2) IWM status in middle childhood and adolescence remains stable independent of changes in maternal sensitivity, and (3) measures of IWMs are more predictive of peer-related social outcomes in childhood than measures of attachment behaviors.

Studies 1 and 2 used data from the nationally representative longitudinal NICHD Study of Early Child Care and Youth Development. Study 1 included 1016 participants who had attachment and maternal sensitivity data at 15-, 24-, and 36-months. Study 2 included 545 participants who had IWM and maternal sensitivity data at third grade, fifth grade, and 15-years. Study 3 was a meta-analysis of 93 studies on the relation between attachment or IWMs and social outcomes with peers.

Results indicated that: (1) A direct relation did not exist between changes in maternal sensitivity and attachment status, but cumulative maternal sensitivity over time was related to attachment status. (2) IWMs in middle childhood and adolescence were stable across a 3- to 4-year interval, but were not stable over a 6-year interval. A relation existed between cumulative maternal sensitivity and IWM status. For the meta-analysis, (3) not enough heterogeneity was found in the sample of effect sizes to merit further analysis. This suggested that measures of attachment and IWMs are not different in their ability to predict peer-related social outcomes. These findings supported the proposal that attachment and IWMs are separate constructs because (a) the IWM in middle childhood and adolescence is more stable than attachment in infancy and early childhood, and (b) early attachment and the IWM in later childhood are differentially influenced by maternal behavior.