Persistence and planning dynamics in sequential effort-based decisions

Sequential effort-based decisions are frequently encountered in our lives. They range from short sequences of actions (e.g., a set of physical exercises in a training session) to repeated long-term efforts required to achieve life goals (e.g., a lifetime of training to become an Olympic champion). However, the way people make sequential effort-based decisions are incompletely known. For example, it remains unclear how fatigue and goal proximity jointly affect persistent behavior toward a goal, and to what extent an initial success in goal pursuit motivates further task engagement. Moreover, it is still unclear how people plan sequences of effort-based decisions, selecting (for example) between investing more effort early on or later—and the possible roles of expertise in selecting these planning strategies. To address these gaps, we designed two experiments investigating complemen- tary aspects of sequential effort-based decision making. In the first experiment, participants performed a repeated effortful task designed to assess the factors de- termining persistence in sequential effort-based decisions. The task consisted of controlling an avatar through a sequence of obstacles until completion, motivated by a reward. A trial ended with the attainment of this reward upon successful task completion. The task varied in both difficulty and length. During a trial, participants could either successfully pass an obstacle, thereby getting closer to the reward, or fail. In case of failure, participants were given the option to continue or to give up on the original task and switch to a less demanding task associated with a smaller reward. The decisions made by participants inform the relative con- tribution of four possible factors—fatigue, goal proximity, early success and lack of progress—to persistence. Our findings indicate that all four factors contribute dis- tinctly to persistence and to its opposite—the decision to "give up". Moreover, our findings suggest that participants consider both retrospective effort and prospective effort when deciding whether to persist or give up. In the second study, we administered a virtual climbing task to compare how expert and novice climbers plan ahead when solving climbing problems of differ- ent difficulty levels. Participants were instructed to report, on a tablet screen, a sequence of optimal hand movements to climb to the top of a wall using a set of available holds. They were expected to accurately simulate the climbing movements in order to identify the best (i.e., least effortful) sequence of actions to solve each problem. We found that the physical difficulty of the problems affected both the speed and the accuracy with which they were solved virtually. Easy problems were solved faster and with fewer errors than difficult problems. Furthermore, our re- sults revealed greater accuracy, speed, and farsightedness in the expert group across difficulty levels. In contrast, novices tended to select more shortsighted (myopic) plans, which reduced short-term movement effort at the expense of the cumulative effort of the overall plan. Their typical mistake was reaching for the closest hold rather than the hold that would be most advantageous for the optimal completion of the sequence. Taken together, these two experiments suggest that factors such as fatigue, lack of progress, goal proximity, early success, and expertise modulate sequential effort-based decisions. These findings highlight the importance of studying effort- based decisions in contextually rich experiments, which allow for the observation of dynamics that unfold across sequences of decisions.