"From Words to Insights: Identifying Themes and Patient Types in Mental Health Care for Older Adults"
Method Stream: Text Mining
ECs: Both 14 and 23 EC thesis possible
Description:
Background
Affective dysregulation in older adults is a complex and often underrecognized issue. Emotional disturbances in this population do not always manifest as classic symptoms of depression or anxiety; instead, they may present through somatic complaints, cognitive decline, or social withdrawal. Assessment and diagnosis are often complicated by age-related factors such as comorbidity, functional decline, or major life changes like bereavement, retirement, or loss of autonomy.
For treatment, the role of an individual’s personal life story—including childhood experiences, trauma, relationship patterns, and life events— is also important in shaping their psychological vulnerability (‘personality’) later in life. These biographical factors often play a central role in understanding why certain individuals develop emotional or psychosomatic problems in old age. In practice, such themes are frequently explored in depth during day-treatment programs, where patients reflect on the meaning and impact of major and minor life events as part of their therapeutic process.
Therefore, gaining insight into both the content of these therapeutic processes and the diversity of patient profiles is essential for improving care and tailoring interventions to individual needs.
Assignment Description
In this assignment, we will conduct a text-mining analysis of around 30 anonymized clinical case files from a specialized day-treatment program for older adults. These patients were referred due to depressive symptoms, anxiety complaints, or predominantly somatic (bodily) symptoms.
The objective of this project is to identify and describe the main themes addressed during treatment—particularly those related to life history, coping, and emotional expression.
Students will formulate a research question, preprocess the textual data, and apply appropriate text-mining methods (e.g., topic modeling, keyword extraction, sentiment analysis). They will interpret their findings through the lens of psychological theory and clinical relevance, with special attention to the role of aging, life experience, and personal meaning.
Who are we looking for?
A student who can work responsibly with patient data and who has an interest in learning and applying text mining methods to real life data. It is important that you understand Dutch as all data are in Dutch!
What do we offer?
This project offers students a unique opportunity to integrate clinical insight with computational methods and to gain deeper understanding of how individual life stories are reflected in the language of mental health care.