Background: Internal Working Models (IWMs) are cognitive structures for simulating and predicting relational outcomes. While self-report attachment scales, such as the Experiences in Close Relationships (ECR), effectively measure the emotional and behavioral dimensions of attachment, they do not directly assess the cognitive function that processes attachment-related information. Transactional Analysis (TA) offers the concept of the Adult ego state, defined as the capacity for reality testing and objective information processing. This study examined whether incorporating a measure of Adult ego state functioning alongside attachment assessment could reveal structural relationships not detected by attachment scales alone. Methods: A total of 31 healthy Japanese adults (7 males, 24 females; mean age = 51.8 years) completed the Experiences in Close Relationships - Generalized Other (ECR-GO) and Tokyo University Egogram (TEG3) before and after participating in a TA-based collage therapy program. Spearman correlations were calculated between ego state scores and attachment dimensions at both time points. Results: At baseline, a significant negative correlation was observed between Adult (A) and Anxiety, that is, negative self-image (ρ = -0.493, p = 0.005), indicating that low self-image and low Adult ego state functioning are associated. This correlation disappeared after the intervention (ρ = -0.312, n.s.), while a new positive correlation emerged between Adapted Child (AC) and Anxiety, that is, negative self-image (ρ = 0.492, p = 0.005). Both Anxiety and Avoidance decreased significantly (p < 0.0001). Conclusions: The inverse relationship between Adult ego state and attachment anxiety at baseline, and the subsequent change in correlation patterns, may suggest structural reorganization consistent with the TA concept of decontamination. This study is a correlation-based analysis and does not constitute causal inference. The concurrent use of cognitive function measures and attachment scales can reveal structural relationships that attachment measures alone cannot detect. Further research with larger samples is warranted.
Hidemi Nakano (Tue,) studied this question.