Health Wearables, Gamification, and Healthful Activity: Fitbit Compliance

Table of Links

Abstract and 1 Introduction

2. Background and 2.1. Leaderboards

3. Effect of Leaderboards on Healthful Physical Activity and 3.1. Competition

3.2. Social Influence

3.3. Moderating Effects of Prior Activity Levels and Leaderboard Size

4. Data and Model

4.1. Data

4.2. Model

5. Estimation and Robustness of the Main Effects of Leaderboards

5.2. Robustness Check for Leaderboard Initiation

5.3. Fitbit Compliance

5.4. Fitbit Attrition, Leaderboard De-Adoption, and Additional Robustness Checks

6. Heterogeneous Effect of Leaderboards

6.1. Heterogeneity by Prior Activity Levels

6.2. Interaction of Leaderboard Size, Rank, and Prior Activity Levels

6.3. Summary of Findings from Heterogeneous Effect Analysis

7. Conclusions and Discussion, Endnotes, and References

5.3. Fitbit Compliance

Step measurement through Fitbit only occurs if the participants wear their devices regularly. We will use the term compliance to refer to the regularity with which a participant wears the Fitbit device. In this subsection, we will probe two compliance-related concerns that may cast doubt on the earlier analyses if left unaddressed. Fortunately, our data include compliance data at a very granular level, which allows us to construct the participant’s compliance measure, percent compliance, at the daily and weekly level, and the participant’s mean compliance for the study duration. We will exploit these data to probe compliance-related concerns.

5.3.1. Do Leaderboards Increase Compliance? The first concern is the possibility that rather than increasing steps, leaderboard adoption increases compliance, which may lead us to observe higher step count purely because of better measurement. To address this concern, we estimated main effects models similar to Equation (1) and the leads-lags models similar to Equation (2) but with daily percentage compliance as the dependent variable and student-day as the unit of analysis. We estimate this model for a number of samples—the entire sample and the subsamples at various mean compliance levels (ranging from 60% to 95%). The main effect model estimates are statistically insignificant and have small magnitudes, ranging from –1.97% to 2.01%. In addition, the leads-lag model’s coefficient plots do not show any sharp increase at or after adoption (see the Online Appendix, Section E.1). These results suggest a null effect of leaderboard adoption on compliance.

5.3.2. Are Leaderboard Effects Discernible at Higher Compliance Levels? The second issue is that some of the participants may have lower compliance and the full sample estimate includes these participants too. Regarding this concern, our empirical analysis would be more convincing if the leaderboard’s effect on participant activity was clearly discernible for participants with high compliance levels. Thus, we estimate impact of leaderboards for participants with high levels of compliance and find these effects to range from 408 to 598 steps (see the Online Appendix, Table E.6). The persistence of leaderboard effects at higher levels of compliance supports the claims from our main results.