Imperfect Inquiry: How Curiosity and Satisfaction Guide Learning

Abstract

Humans are remarkably effective learners, and we are particularly unique in our desire to understand why and how things happen. Why do humans pursue explanations with such fervor? In this dissertation, I explore the phenomenological state that drives learners to pursue explanations (explanation-seeking curiosity) and the phenomenological state that signals inquiry has succeeded (explanatory satisfaction)—the states that together comprise "explanatory phenomenology." In Chapter 1, I present several competing accounts of explanatory phenomenology, which posit different connections between explanatory phenomenology and the goal of learning useful explanations. In Chapters 2-4, I test these accounts. In Chapter 2, I report empirical evidence that explanation-seeking curiosity is related to individuals’ reported expectations that inquiry will lead to useful learning, above and beyond heuristic cues to learning. In Chapter 3, I show that explanation-seeking curiosity is related to expected learning/uncertainty about the query of interest, as operationalized by an optimal model. However, the optimal utility of information and optimal expected learning about a global target of inquiry are less consistently associated with curiosity. In Chapter 4, I report evidence that explanatory satisfaction is associated with perceived learning but is less consistently related to several measures of actual learning. In Chapter 5, I summarize these results and explore their implications for understanding when and why humans experience explanatory phenomenology. Taken together, this work supports the view that explanatory phenomenology is an imperfect motivating force behind useful learning. By experiencing explanatory phenomenology and pursuing inquiry, learners are guided towards asking questions that have strong but fallible potential to lead to informative and useful explanations.