Poster #54 - Parent self-report of state of mind with respect to attachment, infant attachment, and parent sensitivity

• In Event: Poster Session 19

Abstract

Methods.
Sample. 169 12- to 18-month-old infants and their mothers were observed in the Strange Situation procedure (SSP) and a play session coded with the Emotional Availability Scales (EAS).
Procedures. SSP and EAS sessions were scored by trained coders with high levels of inter-rater reliability.
Data on maternal states of mind about attachment were collected using the Berkeley-Leiden Adult Attachment Questionnaire (BLAAQ), providing a measure of mothers’ own childhood experiences and current feelings and ideas about close relationships.
Three international experts on the AAI reviewed 350 items from the BLAAQ (excluding 58 items from the BLAAQ-Unresolved scale) to eliminate items as redundant or theoretically irrelevant. Based on consensus, 137 items were retained to form the BLAAQ Q-set.
The 137 items in the Q-set were sent on cards with instructions for compiling items into four Q-sorts defining reponses of mothers of secure, insecure, resistant, and avoidant infants. A nine-pile forced normal distribution was used (items most salient to the target category at one end and least salient items at the other). Interrater reliability was adequate: Secure (α=.83); Anxious-Avoidant (α=.81); Anxious-Resistant (α =.77).
Results
Q-sort prototype scores. Expert Q-sorts were composited to form 4 sets of prototype item values by averaging scores from each expert’s Q-sorts. Each subject’s pattern of BLAAQ item responses then was correlated with the prototype scores to yield a Q-sort similarity score for each of the 3 attachment categories. The higher the similarity score between responses on the BLAAQ items and the prototype Q-sort, the higher the subject’s score for the category. 145 dyads had SSP and BLAAQ Q-sort data; 120 dyads had EAS plus BLAAQ data.
There was a significant negative correlation between Q-sort similarity scores from Secure and Insecure prototypes (r [145] = -.86, p < .001). The Resistant and Avoidant prototypes were highly negatively correlated (r [145] = -.82, p < .001). Correlations between other prototype pairs did not differ from zero. Thus, adult reports of secure versus insecure strategies were negatively correlated, as were avoidant versus resistant strategies. States of mind specific to avoidance and resistance were not associated with security.
BLAAQ Q-sort similarity scores and infant SSP classifications. Mothers’ BLAAQ responses showed no associations with infant SSP classifications.
BLAAQ Q-sort similarity scores and EAS scores. There were no significant associations between the BLAAQ prototype scores and EAS sensitivity, structuring, and intrusiveness scales. The EAS hostility scale correlated positively with both the Resistant (r [120] = .23, p < .05) and Avoidant (r [120] = .27, p < .01) Q-sort scores.
Conclusions. Mothers’ self-report of states of mind with respect to attachment using BLAAQ prototypes did not predict meaningfully quality of infant attachment with the accuracy and depth offered by the Adult Attachment Interview. Self-report of adult attachment using the BLAAQ may be associated with negative, but not positive qualities of maternal behavior.

Authors

• Mary Josephine Ward, Weill Cornell Medicine
  Presenting Author

• Natasha Stovall