Cognitive performance evaluation under exposure to heat and light color temperature: a bayesian network modeling approach in laboratory conditions

Abstract

Indoor environmental quality significantly affects cognitive performance in educational and occupational settings. While the effects of ambient temperature and correlated color temperature (CCT) have been independently investigated, their mixed effect remains unclear. This experimental study involved 36 students exposed to nine environmental conditions combining three temperatures (22 °C, 35 °C, and 40 °C) and three CCT levels (3000 K, 4000 K, and 6000 K). Cognitive performance was assessed using the 2-back working memory task and the Stroop executive attention task. Two-way repeated-measures ANOVA was used to evaluate interaction effects, and Bayesian network modeling was applied to examine conditional relationships between environmental and cognitive variables. Ambient temperature was strongly associated with variations in reaction time and accuracy (partial η² > 0.60), with increased reaction times and reduced accuracy under exposure to high ambient temperatures. High CCT was associated with modest reductions in reaction tim e , particularly under moderate heat conditions. Bayesian network analysis indicated that ambient temperature had the strongest probabilistic associations with cognitive performance variables within the modeled network, whereas CCT showed weaker dependency patterns. These findings highlight the prominent role of thermal conditions among the environmental variables. The results also demonstrate the potential usefulness of Bayesian network modeling as a complementary analytical framework for exploring multivariate environmental effects on cognitive performance.