Predicting Adherence and Engagement in Web-Based Interventions for Students using Pre-treatment Data
Method Stream: Longitudinal data, intervention
ECs: Both 14 and 23 EC thesis possible
Description:
Web-based interventions have become increasingly popular over the past decades and have demonstrated effectiveness in addressing various mental health disorders and complaints. However, optimizing their impact requires a deeper understanding of the characteristics of individuals before they begin treatment. In particular, little is known about the factors that influence both the likelihood of initiating a web-based intervention and sustained participation. Research on these predictors is often fragmented, with a predominant focus on treatment response rather than the broader process of engagement. It is crucial to distinguish between adherence (remaining in treatment) and engagement (emotional and cognitive investment in the intervention), as fostering engagement may ultimately enhance adherence to treatment and thus treatment outcomes.
This research project will examine data from a comprehensive screening process of over 10,000 university students, identifying predictors of adherence and engagement and comparing profiles of participants who continued in treatment with those who did not. By analyzing clusters of pre-treatment variables, we can gain a better understanding of who is most likely to benefit from internet-based interventions and who might require additional support or alternative interventions.
The focus of this study is on identifying factors that predict adherence and engagement in an web-based intervention, rather than on treatment response. The large screening dataset (N > 10,000) offers opportunities to explore predictors of adherence (e.g., the number of sessions completed), while more detailed engagement parameters (e.g., frequency of logging in, time of session completion) can be examined in the treatment dataset (N=800). The treatment dataset consists of 800 students who entered a three-arm randomized controlled trial (RCT) receiving (1) a human-guided iCBT transdiagnostic program, (2) a computer-guided iCBT transdiagnostic program, or (3) care as usual (CAU) (Koelen et al., 2024). The primary focus will be on using the rich screening dataset and selection process data as predictors and viewing the eventual outcomes of the RCT as secondary.
Research questions that you can focus on in your thesis are (among others):
- Which (combined) factors predict the likelihood of starting treatment?
- Which (combined) factors predict adherence during the early stages of treatment?
- Which (combined) factors predict continued engagement throughout the treatment?
- Are there distinct profiles of students (clusters of predictors) that influence different engagement trajectories?