Virtual reality environments for the study of decision-making processes in risky contexts through the use of physiological measures and behavioural responses

Resumo

Understanding human behaviour in risk situations, how individual and external factors influence our decisions and to what extent it is possible to influence and modify our behaviours, constitutes a challenge both for scientists and for society in general. From the perspective of Occupational Safety and Health (OSH), as well as in numerous fields such as sociology of finance, this topic has important implications since risk situations are a common aspect in various domains of our lives. Risk taking (RT) is part of the decision-making process in uncertain situations, in which the probability of each positive or negative consequence is known in advance. Although the concept of RT is well defined in the literature, it has been approached from different perspectives, so that the factors that have been proposed to explain or moderate RT are also very diverse. Focusing on the individual factors - cognitive and emotional processes - that influence RT, these may affect how hazardous situations are addressed in two different ways. First, they can skew the perception of a situation, so that an adequate evaluation is not carried out and therefore this leads to biased behaviors. Second, these factors shape a certain general propensity towards risk in humans, so that they may or may not be attracted to potentially dangerous situations. In the same way that the definition of RT has been approached from different perspectives, the evaluation of RT has also been treated from different points of view and nowadays constitutes a challenge for researchers and practitioners, so that a clear consensus has not been found regarding the existence of a validated measure for RT. RT evaluation has traditionally been carried out using questionnaires; however, it has been demonstrated that these measures present various limitations that can lead to altered results. Behavioural tasks emerge as an alternative solution capable of overcoming some of these boundaries. Instead, their ability to transference to real life situations appears to be limited. Virtual reality (VR) enables recreating real-simulated situations to carry out performance-based assessments. VR presents numerous advantages that can provide benefits for the evaluation of human behaviours, since it provides greater immersion, fidelity and a higher level of involvement than traditional evaluation methods, and numerous works in the field of applied psychology and organizational neuroscience have endorsed its use for human assessment. In this investigation, we propose VR as technology capable of facilitating the study of RT processes, taking advantage of its numerous possibilities, which can be resumed as: simulation of realistic risk situations, natural interactions with the virtual environment, inclusion of implicit measures for stealth assessment and physiological real-time measurement. This thesis provides novel contributions to the definition of RT, particularly in the identification of which factors constitute this complex process. Moreover, it investigates the use of immersive VR in human behaviour research, specifically for RT assessment, providing design premises of virtual environments for the evaluation of the psychological constructs identified as determinants to define RT. Finally, it analyses the validity of VR in combination with physiological measures for the evaluation of RT in an implicit way.