Modelling Well-Being and Psychological Risk in Doctoral Education: An Integrated Latent Trait Approach

50/2026

Modelling Well-Being and Psychological Risk in Doctoral Education: An Integrated Latent Trait Approach

Friday 19th June 2026

Donnarumma, A.; Guagliardi, O.; Di Stazio F.; Mazza E.; Tanelli M.; Paganoni A.M.

Doctoral education is increasingly recognized as a context in which psychological distress may emerge, shaped by academic demands, institutional environments and interpersonal dynamics. However, evidence on how these latent dimensions jointly configure doctoral well-being remains limited. This study investigates psychophysical well-being, psychological risk indicators and perceived discrimination among PhD candidates at a large public Italian university, using data from an anonymous voluntary questionnaire administered to doctoral researchers on well-being, academic stress, institutional conditions, social support and doctoral experience. Latent constructs of psychological well-being were extracted using Item Response Theory models: among the tested specifications, the Four-Parameter Nested Logistic Regression Model (4PLnRM) provided the best empirical fit, capturing heterogeneity in item response patterns and improving representation of the latent trait structure. Results show that doctoral well-being is primarily driven by personal resources and perceived institutional quality, whereas social support has a comparatively limited association, challenging conventional assumptions regarding the protective role of peer networks in doctoral contexts. With respect to perceived discrimination, a clear asymmetry emerges between vertical and horizontal dynamics. Supervisor-related (“hierarchical”) discrimination is strongly associated with higher stress and poorer psychological outcomes, particularly among women and candidates considering program withdrawal, while peer-related (“horizontal”) discrimination shows weak associations. Overall, findings indicate that doctoral mental health is more strongly associated with supervisory and institutional conditions than with informal support networks, suggesting that improving doctoral well-being may require structural interventions targeting supervisory relationships and institutional governance rather than relying exclusively on individual coping resources.