University of Twente Student Theses


An exploratory study using text mining methods to investigate online therapists' feedback texts for the use of empathy.

Behrendt-Richter, Luis Matthias (2023) An exploratory study using text mining methods to investigate online therapists' feedback texts for the use of empathy.

[img] PDF
Abstract:Introduction: The increasing demand for psychological healthcare services has led to the development of electronic mental health (eMental health) technologies, which provide treatment options for mental health issues via the Internet. However, the extent to which eMental health can contribute to the development of a working alliance between client and therapist is still questioned. This study explored the potential of natural language processing (NLP) methods, specifically text mining, to investigate therapists’ use of empathy in online therapy feedback texts. Methods: A secondary analysis was conducted based on the data from a randomized control trial (RCT), which explored the effectiveness of online-delivered cognitive behaviour therapy for adolescents with insomnia. Online feedback data from seven therapists were analysed using an exploratory text-mining approach. The Orange Data Mining and the Linguistic Inquiry and Word Count (LIWC-22) software were used for investigating the feedback for underlying topics, sentiments, the application of empathy and inter-therapist language style matching. Results: The results of the topic modelling analyses revealed three strongly overlapping topics, all with a functional focus on sleep patterns and advice. The sentiment analyses indicated very similar and predominantly neutral sentiments across the seven therapists. The LIWC-22 analysis suggested a similar level of empathy expressed within the seven corpora of the seven therapists. Finally, the results of the language style matching analysis showed a very high linguistic similarity between all seven therapists, with language style matching scores ranging between 0.89 and 0.98. Discussion: The results of the topic modelling, sentiment analyses and LIWC analyses suggest that the seven therapists in the current study communicated in a very similar, neutral, and functional style concerning their written feedback to their clients and used a similar amount of empathetic language. These similarities in the results could be attributed to the strict protocols within the settings of an RCT study. Therefore, future studies should investigate feedback data from real-world eMental health applications to explore the expression of empathy in a therapeutic context further. The current study suggests that NLP methods provide promising approaches to studying the application of empathy in a therapeutic context and proposes future research at an automated and larger scale.
Item Type:Essay (Master)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:05 communication studies, 17 linguistics and theory of literature, 18 languages and literature, 57 mining engineering, 58 process technology, 70 social sciences in general, 77 psychology
Programme:Psychology MSc (66604)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page