Date of Award

Spring 2016

Degree Name

Bachelor of Science


Economics & Mathematics

First Advisor

Nichole Szembrot


This paper aims to understand whether people follow Bayes’ Rule to update beliefs about important attributes they do not know much about. If they fail to Bayesian update their beliefs, the paper investigates how the external environment, good news versus bad news, risk levels, and the imperfect memory factor contribute to the departure from Bayes Rule. An experimental approach is used to recreate a simplified entrepreneurship scenario in the laboratory environment which was implemented through oTree. 249 participants were recruited through Amazon’s Mechanical Turk. Findings in this paper suggest that people with Low ability have tendency toward overconfidence which persists in all three treatments, under any economic environment, and given good news or bad news. People with High ability, however, have tendency toward underconfidence but become less underconfident under lower risk environment. The surprising result is that people with Medium ability become overconfident under the lower risk environment. Additionally, this paper uses probit models and finds that perfect memory condition and lower risk environment increase the likelihood of taking risk.


Senior thesis completed at Trinity College, Hartford, Connecticut for the degree of Bachelor of Science in Economics and Mathematics.