Intervals, Incentives, and Individual Habits: What Drives Retention in Online Longitudinal Studies?

Longitudinal research on crowdsourced platforms faces substantial attrition, yet there remains little experimental evidence to guide design decisions. We conducted an experiment (N = 1,798) on Prolific, orthogonally manipulating payment rate (£6, £9, or £9/hour plus completion bonus) and session interval (1, 2, or 4 weeks) across five sessions. We also measured habitual tendencies using the Creature of Habit Scale (COHS) and recorded participants’ prior platform experience. Session interval strongly predicted retention: participants in the one-week condition were 1.6x as likely to complete all sessions as those in the four-week condition (80% vs. 50%). Payment rate, by contrast, had no significant effect on retention or response speed. At the individual level, prior platform experience was the strongest predictor of retention, with more experienced participants significantly more likely to complete all sessions. The two COHS subscales showed divergent effects: routine modestly predicted higher retention, whereas automaticity predicted lower retention, challenging the assumption that habit functions as a unitary construct. These findings suggest that researchers designing longitudinal studies on crowdsourced platforms may benefit more from shorter inter-session intervals than from incremental increases in compensation, and that individual differences in habitual tendencies shape attrition in ways that may affect sample composition over time.