JMIR Publications

ABSTRACT

Background:

Contemporary mHealth interventions employ various behavior change techniques to encourage healthier lifestyles. Social comparison stands out as one of the techniques that is consensually agreed to be effective at engaging the general population in mHealth interventions. Yet, it is unclear how this strategy is best employed to engage preadolescents, although they are likely to be concerned with social comparison, since they are particularly developing their social skills.

Objective:

We aim to evaluate how social comparison drives engagement of preadolescents with an mHealth application.

Methods:

We designed a 12-week crossover experiment, in which we studied three approaches to implementing behavior change via social comparison. To leverage naturally existing social structures amongst preadolescents, we hosted this study in their school environment. During the experiment, participants (i.e., both students and teachers) used an mHealth tool that awarded virtual points for performing healthy activities. Participants could read their aggregated score from a leaderboard and compare their performance with others. Particularly these leaderboards were tweaked to implement three flavors of the social comparison technique. The first approach focused on intragroup comparison (i.e., students and teachers competing against each other to obtain the most points), whereas the other two approaches focused on intergroup comparison (i.e., classes of students and their mentoring teachers collaborating to compete against other classes). Additionally, in the third approach, the performance of teachers was highlighted to further increase student’s engagement, through their natural exemplary function. To obtain our results, we used linear modeling techniques to analyze dropout rates and engagement levels for the different approaches. In such analyses, we also considered individual participant traits.

Results:

Our sample included 313 participants (i.e., 290 students and 23 teachers). It was found that students tend to dropout especially in the beginning and during holidays. Also, their engagement levels drop over time and decline during holidays. Still, students do seem to monitor the intergroup competitions more closely than the intragroup setting, since they—on average—checked the mHealth application more often when engaged in team-based comparisons. Students on average performed the most unique activities when engaged in the second intergroup setting (i.e., not highlighting teacher’s performance), perhaps because also their teachers were most active in this setting. Moreover, teachers do seem to have an important role in engaging their students, as their relationship with their students influences how engaged their students will be.

Conclusions:

When employing social comparison to engage preadolescents with an mHealth tool, an intergroup setting—rather than an intragroup competition—motivates them to engage with the application, but not necessarily to perform more activities. It seems that the number of unique activities preadolescents perform depends on the activeness of a role model. Moreover, this effect is amplified by a preadolescents’ perception of closeness to that role model.