PhD defence X. (Xiao) Yu

Dissertation in short:

This dissertation uses different experimental methods to investigate applications of contextualised decisions making. Chapter 2 develops a model-free measurement of attitudes towards intertemporal correlations and tests it in a lab experiment. The findings point out that when it comes to intertemporal risky decisions - that is, decisions concerning streams of outcomes that are both uncertain and delayed - people care about the relations between different time points, which we call intertemporal correlation attitudes. Such attitudes are overlooked in the most popular economic analyses because decisions at each time point are commonly conceptualised as independent from those in any other time point. 

Chapter 3 studies real-world procrastination behaviour in a large-scale governmental program through a series of field experiments. It shows that some costless interventions are effective in alleviating procrastination among a fraction of the general population. This finding connects the wealth of psychology literature on procrastination with another large body of economic literature on real-world applications of nudging. Chapter 4 takes on the prevalent numeracy bias during the COVID-19 pandemics. In an online experiment with a panel structure, it tests the effects of correcting numeracy bias and finds a paradoxical phenomenon called textit{the knowledge-intention gap}. In this case, being exposed to statistical information may lead to cognitive improvements but not to improvements in intentions, or actual behaviour in real life. This finding highlights the importance of a holistic understanding of decision-making mechanisms.