Date of Award


Degree Name

Doctor of Philosophy



First Advisor

Dr. Anthony DeFulio

Second Advisor

Dr. Bradley Huitema

Third Advisor

Dr. Cynthia Pietras

Fourth Advisor

Dr. Wayne Fuqua


Sunk costs are previous investments of time, effort, or money toward a goal that cannot be recovered. People often honor sunk costs by continuing to pursue a goal, despite the availability of an alternative path that would pay off faster, a phenomenon called the sunk cost effect. Prior research has identified variables that influence the sunk cost effect. One variable found in hypothetical scenario-based research and in behavior-based research (Pattison et al., 2011) has been percent of goal completed. The current study was designed to (1) replicate and extend research by Pattison and colleagues and (2) compare results from the same participants on behavior-based and hypothetical scenario-based sunk cost procedures. A pilot study (n = 18) was completed to refine the procedures in advance of the main study. Twenty-five participants were recruited for the main study. Participants completed behavior-based and hypothetical scenario-based tasks to investigate the effects of manipulating percent of task completed on the propensity to continue a course of action. For the behavior-based task, participants were exposed to a sunk cost procedure in which they were presented an initial work requirement of 30 responses. Part of the way through the completion of that requirement they were presented with a choice between staying with that task or switching to another alternative that always required 15 responses. The work requirements were presented in the form of a video game on a computer in which the goal was to kill monsters. A repeated measures design was used to investigate the effect of percent completion of the initial requirement on choice to stay on the initial task. During some trials, it was more beneficial to stay on the initial task (kill the first monster) and during some trials, it was more beneficial to switch to kill the other monster. Staying on the initial work task when it is optimal (requires fewer responses) to switch to the other task is sunk cost behavior. For the hypothetical scenario-based tasks, participants were given a scenario in which continuing on a present course of action entailed losses or was otherwise non-optimal. For example, one scenario entailed completing development of a product that had already been completed by another company. They were asked at various points of completion how likely they were to invest the remaining funds in the project. Overall, participants responded optimally on the behavior-based task and engaged in sunk cost behavior for the hypothetical scenario-based tasks. One explanation for the difference between these tasks is that the behavior-based task may have more discriminable consequences; the consequences for the hypothetical tasks were unknown. Results from the present study, in combination with previous research, suggest that discriminability of a situation is a major determinant of non-optimal persistence on a task. It is possible that what is considered the sunk cost effect in the literature is behavior produced by contingencies in contexts with uncertain or probabilistic outcomes. Verbal behavior may have also played a key role in these results.


Fifth Advisor: Dr. Tracy Zinn

Access Setting

Dissertation-Open Access