Impacting Engagement in Preventive Parenting Programs: A Multiphasic Process

Introduction: Despite evidence that parenting skills interventions can prevent child mental health and substance use problems, participation in these interventions has been notoriously low, diminishing their population-level impact. To address this, prevention scientists must take a theory-based approach to developing strategies that target modifiable predictors of participation. Participation is best conceptualized as an ongoing process of engagement that can change over time. For example, a parent might initially engage by enrolling in a program but later disengage and not attend any sessions. In fact, engagement often drops off dramatically from a parent’s initial interest in participating to program completion. Thus, it is important to delineate different phases of engagement, including interest, enrollment, initiation, retention, home practice, and in-session involvement. It is also important to identify theoretical constructs relevant for each phase, because studies have shown some unique predictors of engagement at different phases. For example, group cohesion has predicted retention but is not relevant for early phases of engagement. This paper presents a theoretical framework relevant to early phases of engagement and uses mediation analyses to test the change theory underlying engagement strategies evaluated in a randomized controlled trial.

Method: The sample (N=122) was drawn from an urban school serving mostly low-income, Mexican American families. Mothers completed in-person interviews at pre-test (53% in Spanish). Families were randomly assigned to an experimental, multi-component engagement package that included a motivational phone interview or an informational brochure plus a non-engagement phone survey. Measures of key theoretical constructs targeted in the engagement package were included in pre- and post-engagement interviews. Parents were offered Triple P, an evidence-based parenting program, which was delivered at the school in Spanish and English.

Results: Tests of mediation provided support for the change theory underlying engagement strategies. Significant mediators of the effect of condition on initiation included: obstacles; mediation effect (ab) =.17 (95% CI = .05, .36), perceived control; ab=.19 (95% CI = .06, .40), and attitudes toward Triple P; ab=.18 (90% CI = .01, .40).

Conclusion: Findings provided evidence that: 1) parental attitudes, perceived obstacles, and perceived control predicted engagement; 2) the engagement package modified these constructs; and 3) modifying these constructs increased initiation. These results demonstrate the value of adopting a theory-based approach to designing and experimentally evaluating engagement strategies to increase the public health impact of effective parenting interventions.