JMIR Publications

ABSTRACT

Background:

Digital self-control tools have emerged as technological interventions to address excessive smartphone usage and promote digital wellbeing. However, these tools face persistent challenges with user attrition and sustained engagement, compromising their long-term effectiveness. Current literature lacks understanding of the psychological mechanisms that drive user retention and meaningful interaction with DSCTs, particularly how intrinsic motivation and readiness to change influence sustained engagement patterns.

Objective:

This study investigates how intrinsic motivation and readiness to change, operationalized through users' daily goal priorities and observable behaviors, can be strategically leveraged to address user attrition in DSCTs and sustain user-nudge interactions to promote digital wellbeing over time.

Methods:

We conducted a quasi-experimental study (n=252) targeting passive users at risk of churning from a DSCT mobile application. Participants were randomly assigned to receive an invitation to reconfigure their nudge settings during daily check-ins (experimental group, n=138) or to a control condition (n=114, no intervention). The experimental group was further classified into acceptance and rejection subgroups based on their response to the intervention. Data collection included system usage logs, self-reported questionnaire responses, and semi-structured user interviews. We analyzed user-nudge interaction ratios, nudge configuration parameters, daily goal selections, and behavioral patterns using t-tests and Cohen's d for effect sizes, at P<.05.

Results:

Of the experimental participants, 46% (63/138) accepted the nudge reconfiguration invitation. The acceptance group showed pre-existing behavioral indicators of higher readiness to change, including 21.53% shorter consecutive usage durations and 20.56% longer cooldown periods compared to the rejection group. Post-intervention, the acceptance group exhibited a temporary surge in user-nudge interaction from 24% to 65%, while the rejection group showed sustained decline below 20%. Behavioral divergence between groups widened significantly (Cohen's d increasing from -0.47 to -0.67, P=.002). Notably, acceptance group participants demonstrated significantly lower tendency to select leisure-oriented daily goals compared to the rejection group (15.6% vs 26.2%, P=.001). Self-reported measures of screen time goals and scrolling regret showed no predictive value for intervention acceptance (P>.1).

Conclusions:

Observable behaviors, rather than stated intentions, effectively predict intervention receptiveness in DSCTs. The study reveals a significant intention-behavior gap, highlighting that behavioral analytics provide superior predictive value compared to self-report measures. Sustainable DSCT engagement requires alignment with users' intrinsic motivation and readiness to change, as evidenced by pre-existing behavioral patterns. These findings suggest that effective DSCT design should incorporate adaptive systems that recognize and respond to users' evolving motivational states while preserving autonomy, rather than relying on static interventions or self-reported preferences.