Date of Degree


Document Type


Degree Name





Arietta Slade

Committee Members

Deidre Anglin

Diana Diamond

Diana Puñales

Steven Tuber

Subject Categories



This study compared maternal reflective functioning (RF) and differentiation-relatedness (DR) during pregnancy and examined how these processes relate to the quality of mother-infant attachment at one year. The subjects were 35 mother-infant pairs drawn from the control group of a longitudinal treatment study, "Minding the Baby (MTB)," a federally and privately funded home intervention program developed jointly by the Yale School of Nursing and Yale Child Study Center, led by Drs. Lois Sadler and Arietta Slade, and targeting a low socio-economic status area of New Haven, CT. The Pregnancy Interview (Slade, 2003) was administered to the women (ages 14-25 years) during the third trimester of pregnancy, and quality of attachment was assessed when infants were 14 months using the Strange Situation (Ainsworth, Blehar, Waters, & Wall, 1978; Main & Solomon, 1990).

The DR scoring system, the Differentiation-Relatedness Scale of Self and Object Representations (Diamond, Blatt, Stayner, & Kaslow, 2011), was adapted for use with the Pregnancy Interview to provide a manual for this study (Daley, 2012). Lowest, highest, and most typical DR ratings were captured for self, the woman's mother, the father of the baby, and the baby. The mean for the baby, at 3.03, was one DR point lower than other relationship means. Three composite scores were created, averaging across relationships: Low DR, High DR, and Overall DR.

Results indicated that maternal RF was correlated with Overall DR and High DR; however, none of these variables distinguished between attachment outcomes. In contrast, Low DR distinguished, with a large effect size (d = .92), between disorganized and secure attachment outcomes (p = .026), and, in post-hoc analyses, between disorganized and all organized outcomes. For the disorganized group, Low DR often dropped to self-other boundary confusion (level 2) across relationships. This suggests that, for a population of women on the lower end of the RF Scale, transient regression to non-differentiated states during pregnancy is a risk factor for disorganized infant attachment outcomes at one year. Results have implications for early identification of high-risk dyads and refinement of intervention models.


Digital reproduction from the UMI microform.

Included in

Psychology Commons