Dimensional Structure of Parent-child Emotion Dialogue: Associations with Internalizing, Externalizing and Trauma Symptoms

Research suggests that emotion socialization, particularly around negative emotions, is a predictor of child socio-emotional functioning (Laible, 2011), also in violence exposed families (Katz & Windecker-Nelson, 2006). Parent-child emotion dialogues may vary along multiple dimensions and these dimensions may differentially relate to child functioning. The Autobiographical Emotional Events Dialogue coding system (AEED, Koren-Karie et al, 2003) describes the quality of parents’ and children’s contribution to emotion dialogues distinct from sensitive/engaged interaction patterns with 7 parent and 7 parallel child scales. The scales reflect concepts of sensitivity and engagement as well as other aspects of dialogue (boundary dissolution, hostility, closure). Usually, the classification is reported (matched-unmatched). No studies have yet reported on the factorial structure of the AEED. Hypotheses 1) the scales of the AEED have a multidimensional structure including dimensions linked with maladaptive parent-child interaction patterns. 2) the found dimensions predict child internalizing, externalizing and trauma symptoms and explain more of their variance than the classification. Study population Two clinical samples of 4-13 year old children exposed to interparental violence (IPV) or sexual abuse (SA) were included. Sample 1: 116 dyads with exposure to IPV, community based, 54% boys, 67% with immigrant background; sample 2: 118 dyads from outpatient children’s trauma centers (exposure to IPV or SA) and a community comparison group, 48% boys. Methods For the AEED task, the experimenter asked each dyad to discuss 4 or 5 events in which the child felt happy, sad, angry, fearful or proud. Dyads were able to carry out the task in their mother language. Transcripts were coded by researchers trained by Koren-Karie. A minority of dyads had a matched classification (24%, 36%). Parents filled out the CBCL (internalizing, externalizing and posttraumatic stress scale (Milo et al., 2013)). Exploratory principal component analyses and regression analyses were performed. Results 1) A four factor solution was found in both samples (Table 2.1, explained variances were 72% and 68%). Two of the factors in each sample were identical: a general sensitivity/engagement pattern (alpha .89/.92) and a closure/resolution of negative feelings pattern (alpha .78/.76). In sample 1, a hostility factor (alpha .64) and a boundary dissolution factor (alpha .67) were identified. In sample 2, two other factors were found, labeled parent and child maladaptive behavior, with child and parent hostility loading on both factors (alpha’s < .50). The two replicated factors were included in regression analyses. 2) After controlling for parents’ education and child age, regression analyses showed no associations between the two cross-sampled replicated factors and child symptoms in sample 1. In sample 2, closure/resolution was associated with less internalizing behavior and the factors together contributed to explaining variance in PTS symptoms. No associations were found for externalizing symptoms. Regression analyses with AEED classification indicated that children in matched dyads showed less CBCL pts symptoms (sample 2). Parents’ and children’s closure of dialogues about negative feelings appears to be a factor distinct from sensitive/engaged interaction patterns. Partial support was found for the predictive validity of these two factors.