Understanding Response Rates in Parenting Studies Using Ecological Momentary Assessment (EMA): Lessons from a Multi-Wave Study

Background and Purpose: Parenting researchers often rely on retrospective surveys to assess parenting practices, but these methods are vulnerable to recall bias and limited by observational constraints. Ecological Momentary Assessment (EMA) offers an innovative solution by capturing parenting behaviors in real time through frequent, brief surveys administered via mobile devices. While EMA reduces recall error and enhances ecological validity, it requires high participant compliance to be effective. Limited research has explored the factors that influence EMA response rates, especially in parenting studies. This study aims to identify participant and study-level characteristics associated with higher EMA response rates, based on data from a multi-wave EMA study on parenting conducted in Central Ohio in 2022.

Methods: Data came from two waves of an EMA parenting study (February–April 2022 and August–December 2022) that included baseline surveys and EMA responses. The analytic sample consisted of 28,681 EMA surveys from 667 parents, of which 93.7% were completed. The outcome variable was binary: whether each EMA survey was completed. We used a multilevel logistic regression model to account for the nesting of EMA surveys within individual participants. Covariates included participant characteristics (age, sex, race, income, education, marital status, number of children, substance use during the previous year, and age of focal child), and survey characteristics (time of day, week of participation, and day of the week).

Results: Adjusting for within-participant clustering, compared to those with only a high school diploma or GED, participants with a four-year college or graduate degree had higher odds of responding (odds ratio [OR] = 1.68, 95% confidence interval [CI]: 1.07-2.66). African American participants had lower odds of response compared to white participants (OR = 0.53, 95% CI: 0.35-0.81). Compared to morning surveys at 10 am, evening surveys at 9 pm were associated with significantly lower odds of response (OR = 0.44, 95% CI: 0.39-0.50). Response rates declined over time, with EMAs in week 1 more likely to be completed than those in week 2 (OR = 1.29, 95% CI: 1.15-1.44).

Conclusions and Implications: This study identified both demographic and design-related factors associated with EMA engagement in parenting research. Response rates were higher among more educated parents and lower among African American participants, indicating potential equity concerns in EMA-based research. Declining response rates in the evenings and across study weeks suggest the influence of survey fatigue, time-of-day burden, or competing responsibilities. Future studies should consider tailoring engagement strategies to support underrepresented parents, such as community-informed recruitment and peer ambassadors. Logistically, shifting evening EMAs to earlier time slots, sending reminders timed to participant routines, and staggered or tiered incentive structures with bonuses for consistent participation could sustain engagement. Additional research might explore factors that influence EMA completion in different demographic groups such as work hours or smartphone or Wi-Fi accessibility. By identifying and addressing the structural and contextual factors that shape EMA response, researchers can design more inclusive, reliable, and representative studies that better capture parenting practices in everyday life.