An Evaluation of the Effects of Decreasing Win Rate on Slot Machine Gambling

Author

Matthew Johnson, Southern Illinois University CarbondaleFollow

Date of Award

12-1-2014

Degree Name

Master of Science

Department

Rehabilitation

First Advisor

Dixon, Mark

Abstract

Animal models can contribute significantly to our understanding of human gambling behavior. However, no proposed animal models of human gambling have been tested using human subjects. The purpose of the present paper was to validate an animal model of gambling with human subjects. Twenty undergraduate rehabilitation students (all women) were recruited and participated for extra course credit. Participants were presented with a concurrent choice between two different simulated slot machines; one machine with symbols and one machine without symbols. During the first 50 choice trials, the payout of the two machines was equated at 50% overall. For the remaining 50 choice trials, probability of winning on the machine with symbols was systematically decreased by 10% overall every ten trials until there was no probability of winning for the last ten trails. On average, participants showed a preference for the machine with symbols during choice trials when win rate was equated; allocating significantly more than 50% of responding to this machine. A repeated measures ANOVA indicated that response allocation to the symbol machine only significantly decreased in the final two conditions (10% and EXT) and did not decrease significantly across any other conditions. Results were also interpreted through behavioral economic analyses. Results indicate that conditioned reinforcement may affect the subjective value of probabilistic reinforcers in humans. These results are similar to those obtained with pigeons under similar conditions and may imply that animal models are relevant to the study of human gambling behavior.