To Bluff or Not to Bluff: Principles and Practice From Laboratory Experiments

Truth-telling behavior in university students (n=38) was examined using a simple stylized poker experiment. Four treatments were applied that changed either card probabilities or payoff structure. Results indicated that in some cases the students’ behavior conformed closely to theoretical predictions. However, other cases indicated robust inconsistencies between observed behavior and mixed-strategy Nash predictions. These findings suggest that there is a strong relationship between the propensity to tell the truth and the costs and rewards thereof. A link between calling behavior and costs and rewards was also found. Implications for the justice system are discussed. Approved: __________________________________________________ Prof. Bill Harbaugh Date