Assessing adult attachment: Relation and validity of two dynamic-maturational model approaches

Abstract

Assessing attachment is essential yet challenging. The Adult Attachment Interview (AAI) provides the best appraisal of adult attachment but is time-intensive and costly. Likewise, concerns have been raised regarding the Berkeley-AAI coding and classification method. Meanwhile, self-report measures of adult attachment are time-efficient and low-cost, but their validity is questionable. The Dynamic-Maturational Model approach to the AAI (DMM-AAI) and a novel self-report measure – the Attachment Relationship Questionaire (ARQ) – may offer a solution. However, additional investigations regarding the validity of DMM-AAI are needed and the ARQ’s psychometric properties have not be tested. The validity of the DMM approach to the AAI and the predictive relationship between the ARQ and DMM-AAI classification were examined for 212 participants living the UK. Results indicated a strong positive relationship between high numbered attachment classification on the DMM-AAI and psychological treatment status, χ²(6) = 56.07, p < .001; Cramer’s V = .371, p < .001. Binomial logistic regressions between the ARQ and DMM-AAI found both single-statement and multi-statement predictive models were statistically significant. However, the ARQ accounted for only a small amount of the variance (R² ≤ 0.15). In conclusion, the DMM-AAI demonstrated strong construct validity, whereas the ARQ is not useful for assessment of adult attachment. Further investigation with a revised version of the ARQ that addresses psychometric concerns is suggested.