Date of Award


Degree Name

Doctor of Philosophy



First Advisor

Dr. Cynthia J. Pietras

Second Advisor

Dr. R. Wayne Fuqua

Third Advisor

Dr. Scott Gaynor


Gambling has been experimentally investigated using various types of gambling simulations designed to mimic the contingencies found in real-world games of chance. Findings from past risky choice research suggest that certain procedures used in existing gambling simulations may systematically increase levels of gambling. Two of these characteristics, the use of a participant stake and the type of options available during gambling, were tested in four experiments in which participants had the opportunity to gamble using tokens exchangeable for entries into a $50 lottery. Experiments 1 and 2 tested persistence on a gamble option when either a single-option or a concurrent gamble no-gamble option was available. In Experiment 1, during concurrent conditions choice of the gamble option probabilistically produced tokens and choice of the nogamble option progressed the game to the next trial. Gambling levels were similarly high in both the gamble and no-gamble options. In Experiment 2, the no-gamble option also produced tokens, i.e., token production, and gambling greatly decreased when token production was concurrently available with gambling. In Experiment 3, a concurrent gamble token-production procedure was used to test preference for a gamble option when participants were or were not staked with tokens prior to a session. Under no-stake conditions, participants could only gamble with tokens earned by choosing the no-gamble option. Choices of the gambling option per choice opportunity were higher under stake than no-stake conditions, but only on the first exposure to the task. Experiment 4 investigated the effects of the value of the tokenproduction option on gambling levels and showed that gambling levels were low regardless of condition. Together, the findings from these experiments suggested that existing gambling procedures that have only a single-gamble option or on which an initial participant stake is given might generate higher levels of gambling compared to when a concurrent token-production option is also available or when participants are not given an initial stake. These findings suggest that common features of laboratory gambling tasks may elevate risk taking and that a concurrent gamble token-production procedure may therefore be a more useful procedure for investigating gambling.

Access Setting

Dissertation-Open Access