Doctor of Philosophy (PhD)


Construction Management

Document Type



Using immersive virtual environment (IVE) based experiments to investigate human thermal states and thermal adaptive behaviors during the building design phase could help identify potential comfort issues and enable designers to optimize building performance cost-effectively before the actual building construction. However, to effectively utilize IVEs to study thermal states/adaptive behaviors, it is crucial to ensure the ecological validity of IVE experiment protocols. This entails accurately representing in-situ (real) environments within IVEs and establishing comparability and generalizability of data collected under different thermal stimuli between IVE and in-situ environments. Existing thermal stimuli based IVE studies have not effectively evaluated the ecological validity of their IVE experimental protocols. Consequently, additional validation is necessary to ensure the reliable use of IVEs in the thermal state domain.

Therefore, this dissertation aimed to further evaluate the ecological validity of the IVE experimental protocol by investigating specific protocol components related to influential factors. First, this dissertation focused on gaining initial evidence on whether an IVE experimental protocol supports reliable thermal state data collection when IVE experiments are performed under different outdoor temperatures. The results of statistical tests showed no association between outdoor temperature variations and thermal states in IVEs, and the IVE experiments created comparable indoor environments despite outdoor temperature variations.

Second, this dissertation aimed to evaluate the ecological validity of the IVE experimental protocol by identifying robust influential factors through their assessment of causation with thermal states/adaptive behavioral intentions. Structural causal models were developed using a causal inference framework to identify robust influential factors causing changes in thermal states/behavioral intentions. The results suggest that indoor room/air temperature and behavioral factors were the robust influential factors causing changes in thermal states and behavioral intentions in both IVE and in-situ experiments.

Finally, this dissertation aimed to guide improvements to the ecological validity of the IVE experimental protocol by investigating the reasons for the observed individual differences in thermal state/behavioral intention responses between IVE and in-situ experiments. The associations between temporal features extracted from influential factors and individual response differences were explored.

Overall, the exploration of several experimental protocol components in this dissertation helped to further evaluate the ecological validity IVEs and highlighted their potential for collecting reliable thermal state/adaptive behavioral intention data during the building design phase.



Committee Chair

Zhu, Yimin



Available for download on Friday, May 15, 2026