PhD Scientific Days 2023

Budapest, 22-23 June 2023

Health Sciences II.

A Unified, Quantitative-Qualitative Method for Exploring the Biopsychosocial Understanding of Health

Dorottya Árva1, Dávid Major1, Anna Jeney2, Annamária Cseh1, Diána Dunai3, Szilvia Zörgő4
1 Department of Public Health, Faculty of Medicine, Semmelweis University, Budapest
2 Division of Global Korean Studies, Graduate School of Korean Studies, Academy of Korean Studies, Seongnam
3 Doctoral School of Sociology, Eötvös Loránd University, Budapest
4 Care and Public Health Research Institute, Faculty of Health, Medicine and Life Sciences, Maastricht University, Maastricht

Text of the abstract

A biopsychosocial understanding of health is key to effective health promotion. In the Balassagyarmat Health Education Program (BEP), medical students enriched their knowledge on health as educators in a near-peer education program. To measure the effects of this intervention on mental models, quantitative methods, limited to predefined variables, were suboptimal.
We wanted to explore the effects of BEP on educators’ idiographic understandings of the biopsychosocial model of health in a systematic manner, and by quantifying, aggregating, and modeling qualitative data across participants, compare BEP educators’ knowledge with other medical students’.
We had two subsamples from Semmelweis University: 1) BEP educators (who learnt all modules and taught in-person), 2) controls (medical students aligned by academic year and sex). We conducted simulation interviews where cognitive task analysis was performed on visual stimuli. We developed codes in a guided inductive process, then quantified data through deductive coding and segmentation with the Reproducible Open Coding Kit.
Code co-occurrences were modeled with Epistemic Network Analysis (ENA) where nodes represent codes, and the thickness of edges captures the relative frequency of co-occurrence between code pairs. Graphs are computed by creating adjacency matrices for each designated segment of data, which are then aggregated by data provider. These cumulative matrices are represented as vectors, normalized, and undergo a dimensional reduction. Vectors are projected into a two-dimensional space where the x axis represents the dimension that explains the most variation in the co-occurrences, while the y axis represents the most variance after the variance explained by the first dimension has been partialled out. With this technique, we can compare the mean networks of subsamples, and of individual data providers.
This unified, quantitative-qualitative method enabled us to investigate the effects of the BEP intervention by aggregating data across participants and comparing subsamples. This is a viable technique for modeling complex relationships without defining variables prior to data collection.
Supported by the ÚNKP-22-3-I New National Excellence Program of the Ministry for Culture and Innovation from the Source of the National Research, Development & Innovation Fund.