Date of Award

Spring 1-1-2013

Document Type


Degree Name

Master of Science (MS)

First Advisor

Matthew Hallowell

Second Advisor

Amy Javernick-Will

Third Advisor

Paul Goodrum


Our study aimed at testing the impact of emotions on construction risk perception in an occupational-like context. To achieve this objective, we induced different positive and negative emotions in 68 Civil, Environmental and Architectural Engineering students at the University of Colorado at Boulder, using validated movie excerpts. Then, subjects’ emotional states were measured using a validated questionnaire. Following the emotional state measurement, participants’ risk perceptions were objectively quantified using a standardized survey embedded within a highly realistic three dimensional virtual environment. This environment, known as SAVES, was created from the building information model of a process plant and over 200 photographs of actual industrial construction projects to ensure a high level of immersion. The measures of emotions served as the independent predictor variables and the risk perception measures served as the dependent variable. Once these data were collected, Principal Component Analysis was used to reduce the dimensions of the predictor variables and differences in risk perception between principal components were tested using the Mann-Whitney U test. The results of these analyses indicated that risk perception differed statistically significantly (p=.009 and p=.003, respectively) between the positive group (happy, amused, joyful, and interested subjects) and the negative group (sad and unhappy subjects), and between the positive group and the self-protective group (fearful, anxious and disgusted subjects). Risk perception differed also greatly between the neutral group and the negative and self-protective groups (p < 0.003). No statistical differences in risk perception were found between the neutral group and the positive group or between the negative and the self-protective groups. iv Interestingly, the negative and the self-protective groups perceived on average almost three times more risk than the neutral and positive groups.