Effects of Ecological Momentary Assessment Prompting Schedule on Affect Measurement Variability and Associations With Next-Day Health Behaviors

Abstract

Ecological momentary assessments (EMAs) effectively measure dynamic changes in constructs such as affect. Accurate assessment of affect variability may depend on study design characteristics. While EMAs delivered at random intervals are typically considered superior for capturing variability compared to EMAs delivered at set timepoints, there have been few experimental comparisons of these assessment methodologies. The present study used data from a nationwide factorial experiment to identify best practices in EMA methods. Participants (N = 205) were randomized to groups across five EMA study design factors, including prompting schedules (random or fixed). Daily EMAs assessed affect and health behaviors for 28 days. Multilevel regressions indicated significantly lower affect variability for the fixed prompt group in 8/16 models: compared to the random prompt group, there was between a 0.18 and 0.36 standard deviation decrease in variability in the fixed prompt group. A significant Prompt × Time interaction indicated that variability in the fixed group decreased over time relative to the random group in 6/16 models. However, prompt schedule did not reliably moderate associations between daily affect means/variability and next-day health behaviors. Thus, differences in affect variability between prompt schedules may exist for some affect items but may not be sufficiently large to influence associations with health behaviors.

Trial Registration: ClinicalTrials.gov number NCT05194228; https://clinicaltrials.gov/study/NCT05194228