Association for Behavior Analysis International

Gambling is a multibillion dollar activity that is associated with simple entertainment, significant financial gains and losses, and damaging compulsive involvement. The variables that influence persistence, rule-following, and subjective discounting in gamblers have been investigated by the authors included in this symposium, through the natural gambling environment (in which people can win or lose personally significant amounts of money) and in experimental settings, using either gambling analogues or hypothetical wagering. These studies indicate that (a) online players of Texas Hold'em persist longer at tables where betting activity is most intense, rather than at tables where they have won or lost the most money (Witts & Lyons), (b) that preferences between slot machines can be influenced by derived rules emerging from prior conditional discrimination training (Wilson & Dixon), and (c) that gains and losses are discounted differentially by participants completing a self-reported functional assessment of gambling (Weatherly & Derenne). The discussant (Ghezzi) will critique these papers from the perspective of having participated in several prior gambling studies using both naturalistic and laboratory settings.

Having access to real gambling data is seldom achievable, given the restrictions of gambling environments. There may be stark differences between contrived gambling situations and those of real gambling situations that may alter the gambling behaviors of individuals. The current study analyzed persistence to gamble in online poker players (N = 10) who gambled in either $3/$6 or $0.01/$0.02 no-limit Texas Hold'em poker. Using poker-analysis software, actual online poker players were tracked over1 to3 days. Data regarding persistence at a table were analyzed with respect to individual wins and losses, table wins, amount of action seen, and whether the player was on one table at a time, or played multiple tables simultaneously. Data were recorded electronically from the commercially available software. Results indicate that persistence is more highly correlated with action, as defined by number of bets won at the table, and the frequency of winning hands that a player experiences, but not with total monetary outcome that the player experiences at the table. These data suggest, then, that players prefer tables with higher pot sizes than tables where individual earnings are high.

The present study investigated whether a verbal rule could be derived and subsequently followed after discrimination training. A multiple-baseline design across6 adult gamblers was used to assess rule stating and rule following during a gambling task. Prior to the gambling tasks, participants were instructed to complete a fill-in-the-blank and a forced-choice worksheet regarding the arbitrary stimuli covering the payout rates on the slot machine (i.e. stimuli A1, A2, and A3). Preference among 2 concurrently available coin options, red and silver, was assessed during the slot machine activity. Duration of play during each gambling task was systematically predetermined, for a total of 24 min. After the first gambling task, a conditional discrimination procedure was used to train3 3-member stimulus classes (i.e., A1-B1-C1, A2-B2-C2, and A3-B3-C3). The C stimulus class incorporated the rule "play red/silver coins." Following discrimination training, rule stating, and rule following were re-assessed. Derived rule stating emerged for all6 participants while derived rule following emerged for 5 of6 participants.

Discounting occurs when the subjective value of a consequence is altered because its delivery is delayed or uncertain. Research on discounting has suggested that how individuals discount is associated with the disorder of pathological gambling. The Gambling Functional Assessment— Revised (GFA-R) is a self-report measure that was designed to determine whether the respondent's gambling behavior is maintained by positive and/or negative reinforcement. Research on the GFA-R has suggested that scoring high on gambling as an escape is associated with the disorder of pathological gambling. The present study was designed to determine whether participants' scores on the GFA-R would be predictive of their discounting of delayed or probabilistic gains or losses. Undergraduate students completed the GFA-R and then a discounting task using the multiple-choice method of measuring discounting. Participants discounted losses differently than they discounted gains and participants' GFA-R scores were predictive of discounting of losses, but not gains. The present results inform us as to how the contingencies that potentially maintain gambling behavior and the process of discounting are related.