Making sense of life review therapy for people with mild depression: human meets computer to obtain additional variables with text mining
How many students possible? 1
Own data collection or existing data? Existing data
Type of research Quantitative empirical
Consider an e-mail or other Internet based form of therapy. It is already possible to analyze the textual interaction between counsellor and client, but what are the most relevant text characteristics to look for when aiming to predict reduction of depressive symptoms? This will be the exploratory research question you will be investigation in an existing dataset of online life-review counselling with people with mild depressive symptoms.
The amount of text data that is nowadays available is rapidly growing. Interest in automated analysis of texts is picking up and psychologists are increasing their involvement in development methods for automatically analyses of texts. Computer scientists and statisticians started this development, but these fields do not have the necessary tradition in the psychological sciences to find the most meaningful characteristics of texts.
Who are we looking for?
A student that considers (or at least not excludes) doing a research master, and who is not afraid to work with computers or statistical analyses.
We do not expect proficiency with any programming language, but we prefer a student who is not afraid to learn some basics of Python or R if necessary.
What do we offer?
We offer three things: if there is a match between the student, thesis and the project, than it would be possible for the student to get involved in a cutting-edge interdisciplinary academic context. Secondly, we offer data where the ideas that are being developed for the thesis can be tested out. Third, if the student is interested in programming we could offer some hands-on experience by working with experts in the several different domains.
Wouter Smink, Anneke Sools and Erik Tjong Kim Sang