Patterns and Family-Level Predictors of Change in Parents' Engagement in a Family Preventive Intervention

Introduction: Beyond attending a sufficient number of sessions, participants in interventions must also be engaged in learning in each session in order to uptake curriculum material and achieve outcomes. For example, participants can show interest in and understanding of curriculum content, actively participate in discussion indicating that they are processing the information, and generally have a positive attitude and interactions with others. Assessing engagement in learning—henceforth, engagement—will allow researchers to discriminate among participants who attended when predicting outcomes of the intervention. Even more fine-grained discrimination may be possible by looking at changes in engagement across sessions. However, it has not been established that engagement varies over time; in previous studies, it has been examined as a trait-like construct. We examine whether engagement in learning in a family preventive intervention systematically changes over the course of the intervention, from the first to last session, and examine the role of parent and family characteristics in predicting level and change in engagement over seven sessions.

Method: 236 parents were randomized to the Strengthening Families Program: For Parents and Youth Ages 10-14 (SFP) or a modified version of the curriculum. Engagement ratings were made by group leaders on 4 items—representing interest, participation, positive affect to leader, and positive affect to other participants—for each parent at each session he/she attended.

Results: We applied a multilevel growth curve analysis to the data using SAS 9.3 proc mixed. We find that engagement does change over the duration of the intervention; the prototypical pattern of engagement in SFP begins high, and changes over time, exhibiting a linear increase and quadratic leveling-off. Parents varied in their initial levels of engagement, and groups of parents who participated together varied in initial levels of engagement and in the linear and quadratic change in engagement over time. Parents in high-tension families showed lower initial engagement. As a trend, on occasions when families had high tension, parents in high-tension families decreased their engagement whereas parents in low-tension families increased their engagement.

Conclusion: Engagement is a dynamic construct. Future work on engagement should recognize that groups of parents may differ in their engagement trajectories, with parents in the same group having more similar trajectories presumably because of shared implementation experiences such as facilitator characteristics or group dynamics. Specific to SFP, efforts to maximize each parent’s engagement should focus on reducing overloads of family tension, which may overwhelm parents’ efforts to change.