Posts, pics, or polls? Which post type generates the greatest engagement in a Facebook physical activity intervention?

Social networking websites have attracted considerable attention as a delivery platform for physical activity interventions. Current evidence highlights a need to enhance user engagement with these interventions to actualize their potential. The purpose of this study was to determine which post type generates the most engagement from participants and whether engagement was related to change in physical activity in an intervention delivered via Facebook. Subgroup analysis of the intervention condition of a randomized controlled trial was conducted. The group moderator posted a new message to the private Facebook group each day of the program. The Facebook posts (n = 118) were categorized into the following types: moderator-initiated running program, multimedia, motivational, opinion polls, or discussion question and participant-initiated experience shares, or questions. Four metrics were used to measure volume of engagement with each post type, "likes," "comments," "poll votes," and "photo uploads." One-way ANOVA was used to determine whether engagement differed by post type and an independent samples t-test to determine differences in engagement between moderator and participant-initiated posts. Pearson correlation was used to examine associations between total engagement and change in physical activity. Engagement varied by post type. Polls elicited the greatest engagement (p ≤ .01). The most common form of engagement was "likes," and engagement was higher for moderator-initiated rather than participant-initiated posts (mean = 8.0 [SD 6.8] vs. 5.3 [SD 3.2]; p ≤ .01). Total engagement with the Facebook group was not directly associated with change in physical activity (r = -.13, p = .47). However, engagement was associated with compliance with the running program (r = .37, p = .04) and there was a nonsignificant positive association between compliance and change in physical activity (r = .32, p = .08). Posts requiring a simple response generated the most engagement. Intervention moderators should facilitate familiarity between participants at the intervention outset, to encourage engagement between participants. Engagement was related to change in physical activity, and these recommendations should be incorporated to enhance engagement and efficacy of interventions.