Inattention, task completion, and retention among undergrads

Three economic experiments are used to answer questions about human decision making: when to pay attention, when to complete a task, and when to enroll in second year university. Chapter 1 explores inattention as an explanation for economic mistakes like overpaying. A novel experiment design separates intention to collect information from its actual collection and use. The results show that many mistakes are the result of optimal inattention, when someone is inattentive because the costs of information exceed the benefits of paying attention. But the experiment also demonstrates addressable mistakes which are driven by dynamic inconsistency and ineffective attention, even in a very simple information setting. Chapter 2 explores procrastination through a novel real-effort task experiment with no commitment. Unlike many previous economic experiments on intertemporal choice which use monetary rewards or record choices at a single point in time, this experiment finds participants have a preference to complete the real-effort task early, in line with psychological experiments on working memory. Structural estimation of the classic model of hyperbolic discounting finds the present-bias parameter $\beta > 1$, indicating a future-bias and desire to finish costly tasks immediately. Chapter 3 uses a novel control group in a field experiment to study the causal effect of a first-year seminar on undergraduate student outcomes such as enrollment, academic standing, GPA, and psychological well-being. Students enrolled in seminars have higher GPA and retention than the overall pool of students, but these effects disappear when comparing seminar students to the novel control group of students who were willing but not enrolled in a seminar. Self-selection into a seminar course may be indicative of student learning style and strengths in ways that are not measured by registrar data; this experiment demonstrates that studies that fail to control for willingness to enroll are at risk of identifying selection effects rather than genuine effects of a seminar. The participants sampled in all chapters are drawn from the same undergraduate population, though they are distinct.