JMIR Publications

ABSTRACT

Background:

Facebook is a suitable platform for public health interventions. Facebook users can express their health behavior in many ways using interaction buttons. These interactions may be associated with the users’ motivations for behavioral change and may affect the reach and engagement of public health interventions.

Objective:

This research aimed at understanding how Facebook users' interactions correlate with organic reach, engagement, and motivational language regarding to the same smoking cessation support contents.

Methods:

The study population was the audience of a public Facebook page. We included 1025 smoking cessation support Facebook posts (N=1025) which were targeted at smokers. The following data were collected from the 'Post Details': organic reach, the number of each ‘engagement’ interaction (like, love, haha, wow, sad, angry reactions, shares, comments, clicks), and the total number of negative interactions. Furthermore, comments were analyzed using the motivational interviewing approach to distinguish major and minor categories of smokers' motivational language. Overall, these data were compared using Spearman correlation.

Results:

Surprisingly, we found a significant negative correlation between organic reach and 'like' reaction (rs=-.418; p<.001). The strongest significant positive correlations of organic reach were observed with 'haha' reaction (rs=.396; p<.001), comments (rs=.368; p<.001) and clicks (rs=.300; p<.001). Contrary to expectations, the interactions of engagement, especially ‘like’ reaction, were sharply separated by significant negative correlations from 'wow' (rs=-.077; p=.013), 'sad' (rs=-.120; p<.001), 'angry' reactions (rs=-.136; p<.001), and comments (rs=-.130; p<.001). Additionally, a high number of negative interactions were significantly associated with 'wow' (rs=.076; p=.016) and 'sad' reactions (rs=.091; p=.003). Finally, some specific Facebook reactions or their combination were significantly correlated with major and specific minor categories of motivational language. For example, significantly positive correlations were detected between 'love' reaction and 'change talk need' phrases (rs=.089; p=.004) or 'angry' reaction and 'change talk need' phrases (rs=.068; p=.029).

Conclusions:

A potential practical implication of our results is to assist public health professionals who design Facebook-based interventions in increasing organic reach and engagement. Suggestions for giving ’haha’ reaction, comments and clicks, rather than 'share' and 'like' interactions, should be made in order to get high reach. Negative reactions and negative comments should be ignored in assessing engagement. Some interactions can specifically predict the motivational language of behavioral change. Regarding our findings, we suggest implementing continuous evaluation of Facebook interactions during public health interventions.