Detailed contribution information
| Contribution title | Major life events from adolescence to young adulthood: A longitudinal natural language processing analysis of a large urban cohort |
|---|---|
| Contribution code | D1.136 |
| Authors | |
| Form of presentation | Poster |
| Topic |
|
| Abstract |
Background: Recent research shows an increase in mental health problems among young people oftentimes labelled 'youth mental health crisis'. Even though youth mental health is a growing area of concern, much of the research is based on predefined survey instruments. This limits the understanding of the subjective experiences and evolving priorities of youth. Aims: Our study addresses this gap by integrating youths‘ first-hand accounts using innovative Natural Language Processing (NLP) techniques to uncover risk and protective factors. Specifically, we aim to investigate openly assessed major life-events in youths and describe how key life-event topics change from mid adolescence to early adulthood. Methods: In the Zurich Project on Social Development from Childhood to Adulthood (z-proso), 1,442 participants answered a single-item open-ended question on their most significant life-event in the previous years at four measurement occasions between the ages 15 to 24. We analyzed themes in N=5,708 text narratives using topic modelling with the Python library 'BERTopic', combining conventional techniques with large language models (LLMs) and analyzed shifts in life-event topics over time. Results and Conclusion: Results display a diverse range of youths’ life-events across multiple life domains (education & career development; social relationships, leisure activities & successes; mental health & well-being; and life-events related to other transitions & independence). Most life-events were of positive valence (83.2%). Major life-event topics showed distinct developmental shifts over time. Our work, thus, highlights salient life-event topics across different ages and illustrates how longitudinal population-based research can draw on text data through NLP techniques to assess lived experiences of youth. |