Decision‐making styles and mental health—A person‐oriented approach through clustering

Abstract

Previous research examining the associations between decision-making styles and mental health has neglected to look at the configurations of styles. The study identifies mutually exclusive groups of people according to their combination of the five decision-making styles and examines the differences between them in mental healthrelated variables in two studies. In Study 1, decision-making styles, mental health indicators (well-being, depression, stress) and socio-psychological protective factors (resilience, optimism, social support) were examined in university students in two rounds 1 year apart. A cluster analysis revealed three distinct decision profilesindependent/nonintuitive, avoidant/spontaneous and rational/nonavoidant. The best values of mental health-related variables were found in the rational/nonavoidant profile and the most negative values in the avoidant/spontaneous profile. The decisionmaking styles were stable after 1 year although they were not able to explain changes in mental health-related variables. Study 2 identified three similar clusters in a more general sample-spontaneous/irrational, dependent/avoidant and rational/nonavoidant-with the lowest level of psychological distress symptoms in the rational/nonavoidant cluster. The study highlights the usefulness of combinations of decision-making styles with regard to mental health as well as noting their specificity according to sample characteristics.

Machine learning identifies a COVID-19-specific phenotype in university students using a mental health app

“…K-means clustering was used to characterize participants, identify the degree of heterogeneity within the population, and to identify how many clusters there were in each timepoint. The k-means clustering algorithm, specifically, was selected as it has been previously used by studies examining the mental health of university students ( Di Benedetto et al, 2019 ; Bavolar and Bacikova-Sleskova, 2020 ; Liu, 2021 ; Nelsen et al, 2021 ). The optimal number of the k-means clusters was determined using a Silhouette score and distance metric chosen was Euclidean distance.…”

Achievement Profiles in Math, Science, and English: Exploring Contextualized Sex Differences

“…This is difficult to achieve in variablecentered studies, where inferences are made on single, or a set of, variables (e.g., math achievement scores alone or science achievement scores alone). The value of determining person profiles from variables have been demonstrated in various research (e.g., Bavolar & Bacikova Sleskova, 2020;García Mendoza et al, 2019;Heikkilä et al, 2011;Lyndon et al, 2017;Vettori et al, 2022).…”