JMIR Publications

ABSTRACT

Background:

Smartphone apps to promote physical activity (PA) are abundant, but few lead to substantial and sustained behavior change. Many PA apps purport to induce social comparison processes by prompting users to compare themselves to others, as this process has shown benefits for PA and other health behaviors. Although social comparison may motivate PA for some users under some circumstances, two important areas remain unclear: (1) how comparison processes are defined or harnessed in existing PA apps, and (2) to what extent advances in the field social comparison are applied in these apps (e.g., consideration of demonstrated between- and within-person variability in response to comparison opportunities), which may be useful for tailoring comparison features.

Objective:

Toward the goal of improving social comparison features of PA apps, the objective of this meta-review was to summarize existing reviews of behavior change techniques in PA apps, with respect to app features categorized as inducing social comparison.

Methods:

The authors searched PubMed, Web of Science, and PsycINFO for reviews of PA smartphone apps. Of the 3,743 initial articles returned, 26 reviews met inclusion criteria. Two independent raters extracted data from these reviews, including the definition of social comparison used to categorize app features, the percent of apps categorized as inducing comparison, specific features intended to induce comparison, and any reference to tailoring comparison features. For reference, these data also were extracted for related processes (behavioral modeling, norm referencing, and social networking).

Results:

Eight review articles (31%) categorized app features as prompting social comparison; the majority of these employed Abraham & Michie’s [1] definition of comparison, which differs from definitions in later iterations of the same taxonomy. Very few reviews specified what users were intended to compare (e.g., steps, physical fitness) or which features were categorized as inducing comparison (e.g., leaderboards, message boards), and no review referenced tailoring of comparison features. In contrast, 14 reviews (54%) categorized features as prompting behavioral modeling and 8 (31%) referenced tailoring app features to users’ personal goals or preferences.

Conclusions:

The observed heterogeneity across existing reviews of PA apps and the lack of critical information (e.g., which dimensions or features were relevant for comparison) leaves room for confusion about how to harness social comparison to increase PA. This confusion may preclude strong conclusions about the effectiveness of social comparison app features in future research. Further, this review generated no evidence that existing work has applied important findings from the broader social comparison literature, which shows that users have differing preferences for and responses to social comparison information. Greater integration of these findings (both between- and within-person) may increase the effectiveness of social comparison features of PA apps, thereby improving the public health impact of these mobile health tools. Clinical Trial: N/A