Evaluating the RD Parent Empowerment Program: A pragmatic approach within real-world constraints

Background: Concern about the national crisis of poor diet and physical inactivity sparks strong interest in local interventions. Evaluating and disseminating results of such interventions is vital for informing future practice and maximizing funding efficiency. Yet evaluation is hampered by substantive challenges, including: limited budgets, participant burden, and measurement and IRB issues. This presentation describes a pragmatic evaluation approach developed within these challenges, for a community-based family nutrition initiative.

Methods: The RD Parent Empowerment Program was a workshop series promoting healthy family food and activity practices, implemented in lower-income neighborhoods in 4 cities. Funding was primarily for program delivery; to make evaluation feasible from the outset, we adopted guiding principles: efficiently use limited resources (funding and participants’ time); maximize validity using multiple methods; and focus on key intervention goals and targets.

Results: Applying these principles, we developed a mixed-methods evaluation, combining quantitative pre-post data generated within the intervention itself (self-assessments of family environment and practices), with qualitative data from focus groups with parents and leaders, conducted by experienced local moderators after programming ended. Pre-post self-assessment scores improved significantly, with corroborative evidence provided in both parent and leader focus groups.

Conclusions: The pragmatic evaluation design yielded informative results. Although not without limitations, this may offer a feasible approach to evaluating local programs and disseminating results absent research/extensive evaluation funds. Moreover, external validity may be a strength of this approach. Refining such pragmatic designs may facilitate more widespread program evaluation, and advance discussions about the nature of evidence demanded for effective programming.