Post-reinforcement pause was examined to determine the reinforcing value of a win, loss, and a loss disguised as a win (LDW) for gamblers at multi-line video slot machines. The study was conducted in naturalistic settings across a variety of participants, age 21 years and older. The length of the post-reinforcement pause was recorded using a stopwatch for one win, one loss, and one LDW for each participant and was measured by recording the time between the outcome delivery and the initiation of the next spin. The different times were evaluated to determine which of the three resulted in the longest post-reinforcement pause for the gamblers following the slot machine outcome. The present study replicates and extends previous research on post-reinforcement pause in slot machine gambling, and provides discussion around the clinical utility of such findings on the prevention of problem gambling. Problem gambling is an epidemic, and there are numerous variables that contribute to its development. Post-reinforcement pause is one for those factors, and the present study can help us gain a better understanding of the events that maintain problem gambling and ways to prevent it. The results of the present study found that wins are the most reinforcing to gamblers compared to LDWs and losses, and that LDWs are significantly more reinforcing to gamblers than losses, as indicated by the patterns of the post-reinforcement pause.

California Quail were exposed to fixed-interval schedules whose values ranged from FI 90 to FI 180. Post-reinforcement pause lengths and terminal rates were recorded and grouped into relative frequency distributions. The relative post-reinforcement pause length was found to be an increasing function of FI value such that at larger FI values a proportionally greater period of time was taken up by the post-reinforcement pause. Terminal rate (rate during the final 20% of the interval) was a decreasing function of FI value. The large amounts of variability in terminal rates observed indicated that terminal rate in fixed-interval schedules is not constant from interval to interval as is often reported in the literature. For a given subject, when overall rate of response for a session was plotted as a function of mean pause length for that session, no consistent relationship was found. Among the subjects there were two to three-fold differences in overall rate on FI 90, the only value to which all subjects were exposed. Differences among subjects in mean overall rates were correlated with differences in mean pause lengths, however. Thus, a subject's performance on a fixed-interval schedule could be characterized in terms of pause length and overall rate although rate in any given session was not necessarily correlated with mean pause length for that session, Differences between subjects in mean overall rates were also correlated with differences in mean terminal rates.