Human beings perform actions for desirable outcomes to satisfy needs and experience pleasure. Such behavior can be cued by the environment. The goal-directed account for such cue-based behavior has been well-accepted in more recent studies, especially in the field of automatic goal pursuit. However, the understanding and examination of the direct control of goal-directed behavior by the environment are plagued by the absence of clear methods or tests. More specifically, while studies hint at the involvement of goal representations, they need to properly distinguish between behavior activated directly by the cues and behavior mediated by representations of desired outcomes. That is, the cues used in these studies are usually associated with both the goal and the instrumental actions causing them in daily life. Therefore, it is impossible to distinguish between the direct and indirect effects of cues on behavior.
To address this issue, in the present thesis, a Pavlovian-to-Instrumental Transfer (PIT) paradigm was used to systematically investigate cue-based goal-directed behavior by comparing cued behavior toward high vs. low-value outcomes. This paradigm allows researchers to separate action-outcome learning from stimulus-outcome learning, meaning that the cue is only indirectly associated with actions that lead to the same outcomes. Hence, any facilitating effect from cue can only be attributed to the indirect effect of outcome representations. More importantly, we investigated such effects on goals with social meaningfulness (e.g., pro-social goals) and what stage of behavior (i.e., action initiation and action persistence) can be influenced by such processes.
Results indicate that cues predictive of low or high-valued outcomes prompt instrumental responses when the cue and response share the same outcome. Moreover, the effects are more pronounced for high-value outcomes, indicating a value-based specific PIT effect. For the high-level goal pursuit, we found that a cue predictive of pro-self outcomes facilitates instrumental responses, whereas the specific PIT effect for pro-social outcomes only emerges when participants have the freedom to donate the money. Additionally, we examined the influence of multi-functional outcomes on the PIT effect. Results reveal that cues associated with multi-functional snacks facilitate the corresponding actions, whereas cues associated with single-functional snacks do not. Furthermore, we explored the motivational control effect of cues on goal-directed behavior. The findings demonstrate that participants responded faster to the high-value cue. However, this effect was only found in action initiation and did not translate into action persistence over time.
Overall, this research sheds light on the environmental cues that influence human behavior, highlighting the mediating role of outcome value and providing insights into high-level goal pursuit cued by the environment, at least in action initiation. We believe that the PIT paradigm is a valuable tool for investigating the motivational underpinnings of cue-driven goal-directed behavior and, thus, how automatic goal pursuit may emerge.