Title: Mining for meaning in e-health
Type of assignment: Ma
Internal– or external?: Internal
Number of students: 2
New data collection: no
Type of research: quantitative (text mining; advanced statistics)
Nr of EC: 10 or 30 EC
There is increasing evidence that e-health interventions are equally effective for reducing depressive symptoms and increasing well-being as face-to-face interventions. However, existing research does not explain huge differences in effectiveness between individuals. Why do some individuals benefit more from therapy than others? And why are different treatments equally effective? To answer these questions, more insight is needed in the processes leading to therapy success. To analyze the processes involved in e-mail guided counselling, the research project ‘What Works When for Whom’ aims to employ text mining methods (e.g. automated text analysis).
In this assignment, you will participate in this project and contribute to the development of the exciting new field of text mining in psychology. Your will explore methods for automatic (computerized) analysis of e-mails between counsellor and client. The type of interventions involved are based on positive psychology, acceptance-and-commitment-based therapy, life review, and expressive writing. You will use existing software for the automatic detection of text-based patterns (e.g. content-based by investigating cognition- and emotion-related words, or grammar-based by investigating the use of pronouns, active/passive sentence constructions), and relate these to outcome measures (reduced depression and increased well-being).
Possible research questions include:
1.Which text-patterns discriminate most optimally between successful and unsuccessful treatments?
2.Which text-patterns discriminate most optimally between successful and unsuccessful counsellors?
Supervisors: Anneke Sools & Wouter Smink