Development and application of an experimental framework for the use of virtual reality to assess building users’ productivity, comfort, and adaptive-behaviour

The use of Virtual Reality (VR) to assess comfort, productivity, and behaviour in built environments is currently emerging. However, an effective and standardised methodology to support researchers in performing VR tests to collect reliable data is lacking. Thus, the study aims to develop a novel and comprehensive experimental framework for organising the application of VR in this research domain, based on five validity types: content, internal, face, ecological, and criterion validity. To illustrate its suitability and applicability, the framework was applied to a singledomain case study of a virtual office room. 52 volunteers were recruited to perform cognitive tests (Stroop test, OSPAN test, Magnitude-Parity test) and answer questionnaires. Tests and surveys have been developed to support the content validity of the research. Each test was performed under different thermal stimuli: 24 °C and 16 °C, randomly assigned, to pursue internal validity. The first goal was to demonstrate that the framework allowed the creation of highly immersive scenarios. Findings confirmed the ecological validity of the model with an excellent sense of presence, graphical satisfaction, involvement and realism and non- significant levels of cybersickness. The second aim concerned the evaluation of the ability of the VR environment to capture the influence of the temperature set point on the dependent variables. As expected, a statistically significant influence was detected only on thermal comfort votes and adaptive behaviour, thus supporting the criterion validity. Results highlighted the values and potentialities of applying the present framework in the context of the emerging multi-domain research in support of user-centred design.