Studying Mechanisms of Change in a Trauma-Focused Parenting Intervention: Using Qualitative Data to Understand Quantitative Measurement

Background: Exposure to illicit substances in utero is associated with risks for children’s cognitive, social and emotional development (Salo & Flykt, 2013). Additionally, relationships with caregivers and caregiving practices are associated with developmental outcomes for these children (Nair, et al, 2003).  Substance dependent mothers often have difficulties regulating their emotions and attuning to their infant’s needs given histories of trauma and mental health disorders. Their children may suffer maltreatment and the attachment relationship can be strained (Salo & Flykt, 2013). Reflective functioning (RF), which refers to a parent’s ability to understand a child’s behavior in terms of thoughts, feelings and intentions, is associated with a child’s attachment security and development (Slade, 2005). Some parenting interventions offered by social workers have sought to enhance RF as a way to improve caregiver-child interactions, thereby minimizing the developmental and maltreatment risks for the child (Paris, et al, 2015). This study seeks to understand change in RF by using qualitative data to understand a quantitative self-report measure.

 

Method: Sixty-six mothers in treatment for primarily opioid dependence who participated in a dyadic, attachment-based parenting intervention completed the newly developed Parental Reflective Functioning Questionnaire (PRFQ, Luyten, et al., 2009) pre- and post-treatment, as well as semi-structured post-treatment interviews detailing parenting and intervention experiences. Most mothers were Caucasian, unemployed, in their late 20s/early 30s with extensive trauma histories and symptoms of psychological distress. Their children were on average 26 months old and had already experienced multiple traumatic events. Interview transcripts were initially coded broadly for RF by 2 RAs supervised by the PI using grounded theory techniques (Charmaz, 2006). Subsequently, post-treatment PRFQ scores on the three 7-point subscales, pre-mentalizing (PM-low is ideal), certainty of mental states (CMS-middle is ideal), and interest and curiosity (IC-high is ideal) were used to guide re-coding of the RF interview segments to compare the measurement findings with descriptions provided by the participants themselves.

Results: Forty-six of the 66 participants described RF processes in their interviews. Those scoring lowest on PM described an increasing awareness of their child’s complex behavior, no longer ascribing it to misattributions such as considering it a “conspiracy” against them. Participants who perceived a shift in their CMS described greater uncertainty in “knowing what [the child] needed” but felt hopeful that they could “figure it out a bit quicker.” Mothers who endorsed high IC at post-treatment described improved curiosity about  their child’s behavior including “what she was trying to communicate when she was upset,” plus a desire to “put myself into their shoes” in order to understand emotions and behaviors.

 

Conclusion: Qualitative findings served to clarify the meaning of a newly developed quantitative measure of RF. Given the intervention’s aim was to improve parenting capacities, evidence provided by two sources of data strengthens the argument for possible intervention effectiveness. Reliance on multiple methods to understand change mechanisms enriches our understanding and furthers development of effective evidence-based social work practices.