A Mixed-Methods Approach to Exploring Engagement in MoodTech: An Online CBT Intervention for Older Adults with Depression
In recent years, online cognitive behavioral therapy (CBT) interventions have played an increasing role in treating late-life depression. Previous studies have reported that online CBT interventions can be effective in treating depression and promoting behavioral changes among older participants. However, inadequate engagement could potentially weaken their effectiveness, and the motivations and barriers to participation among older participants in these interventions are less well known. This study aimed to obtain deeper understandings of older participants’ engagement patterns and subjective experiences of MoodTech, an online CBT intervention tailored for older adults with depression. I employed three different methods to analyze diverse forms of data produced during the intervention. There were three aims. First, I characterized the engagement of participants through visual analysis of log data. I visualized the frequencies of their online activities and skills practices, then identified patterns in each measure of engagement. A meta-pattern graph was created to facilitate identification of groups of individuals who shared similar patterns across engagement measures. Second, I conducted a network analysis of the participants who had access to the peer interaction features and compared engagement behaviors in three kinds of peer interaction networks (comments, likes and nudges). Third, I performed a qualitative analysis of the textual data, including messages, posts, comments and thought records of the participants. Using a qualitative analytic method based on Grounded Theory, I examined the application of CBT and how participants responded to CBT and engaged with the intervention. I also explored potential explanations for the observed behaviors in individual and network engagement patterns. With regard to the results, for the first aim, I observed great diversity, but also similarities, in patterns of engagement among participants. For the second aim, I found that the networks of nudges were less dense than the comments and likes networks and there were fewer people involved, which may show that older people attached importance to the actual contents of interactions. Last, I found evidence from the qualitative analysis that many participants learned CBT strategies and practiced them to understand why they were having sustained feelings of depression and low productivity. Some even successfully broke harmful cognitive or behavioral patterns. But other older adults encountered obstacles due to shortcomings in the website design or were reluctant to practice the CBT strategies because of previous unsuccessful experiences with CBT. Overall, engagement behaviors of older adults in online CBT interventions are hard to predict, but can potentially be influenced by technology use habits, contents of social interactions, and previous psychotherapy experiences. Future intervention design may take these findings into consideration and adjust lesson contents for different subgroups of participants, as well as improve the usability of the intervention to meet the needs of older adults. The methodology of this study also shows that combining multiple methods is feasible and provides a richer characterization of engagement from different perspectives.