: Change in Family Systems: Tracking Family Functioning through Intensive Daily Data Collected during Interventions

Introduction: Intensive daily data collected during an intervention can elucidate the course of change between pre and post, revealing when outcomes change, the shape of change, etc. Because the utility of this approach to program evaluation rests on there being change, fluctuation, and/or variability among participants during the course of an intervention, the first step in this research is to describe these aspects of the data. In family interventions, fluctuation will be especially important; family systems theory predicts that attempts to adapt to a disruption of the system (manifested as fluctuation, day-to-day variability) precedes system change and reorganization (Minuchin & Fishman, 1981). Therefore, we expect that families will experience a “messy” period of adaptation while they attend the intervention, which is our measurement time frame; specifically, parents who attended more sessions will show greater fluctuation.

Methods: Participants were parents randomized to intervention conditions in a trial with the Strengthening Families Program: For Parents and Youth 10-14 (SFP). Parents completed daily web-based surveys that assessed select outcomes 6 out of 7 days of each week of the intervention. Outcomes included: SFP-targeted parenting behaviors (3 items, support, guidance, anger management), mindful parenting (5 items), parent-youth affective quality (4 items), and youth-parent affective quality (4 items). All scales are averages of items rated from 0 to 100.

Results: We will model outcomes using multilevel models to find an appropriate fit to the data for each outcome. Models show that there is not a significant prototypical pattern of linear or quadratic change in outcomes during the intervention. The data does show precursors to system change; there is a large degree of day-to-day variability (e.g., 55.78% within-person variability, parent-youth affective quality) in parents’ reports of outcomes. To explore this variability, we will examine models of heterogeneous variance, using attendance/dosage to predict differences among parents in the day-to-day variability of their reports. Analyses indicate that participants with higher attendance report more variable outcome scores from day to day (e.g., tau = 0.19, z = 4.47, p < .01, parent-youth affective quality).

Conclusions: These analyses will answer the question of whether dosage predicts the degree of disruption—whether the intervention is initiating change in the family system. In future analyses, participants reporting high disruption/variability will likely be the participants who experience the most pre-post change from the intervention. This data suggests that, in response to intervention, families follow a dosage-disruption-change sequence.