Abstract
In windowless enclosed workspaces that rely entirely on electric lighting, selecting appropriate lighting parameters is essential for evaluating the luminous environment. Light can influence multidimensional human responses through both visual and non-visual pathways. This study compared conventional and non-visual lighting parameters using data from ten participants in a nine-day daylight-free experiment under five lighting conditions. Thirteen cognitive, oculomotor, and autonomic responses were analysed with an illuminance × correlated colour temperature (CCT) factorial model and four single-parameter linear mixed-effects models for illuminance, CCT, melanopic equivalent daylight illuminance (mEDI), and circadian stimulus (CS). Before false discovery rate (FDR) correction, visual task performance showed illuminance-related associations and psychomotor vigilance task (PVT) median reaction time showed associations with CCT and non-visual parameters; these behavioural findings did not remain significant after correction. Oculomotor measures showed the most robust associations, with multiple effects of illuminance, mEDI, and CS remaining significant after FDR correction. No autonomic association survived multiplicity control. Overall, different lighting parameters were associated with different types of human responses. These findings indicate that no single parameter can characterise all measured human responses. The results provide an experimental basis for selecting lighting parameters according to task requirements and for optimising lighting design in enclosed workspaces.