Repository for a working paper by Linus Hof , Veronika Zilker , and Thorsten Pachur .

In many situations, we are initially lacking information about our choice options. Making a decision therefore not only requires a procedure for comparing options, but also a procedure for searching for samples of information and stopping search. Although search, comparison, and stopping are essential building blocks of most decision-making processes, their interplay has not yet been systematically studied, let alone formalized in a computational model. In this article, we develop a general formal framework to specify sampling strategies for experience-based risky choice in terms of a search, a comparison, and a stopping rule, and examine the effects of such rules and their interaction on risky choice. Our analyses demonstrate how descriptive hallmarks of decision making under risk—deviations from maximization, risk aversion, and over- or underweighting of rare events—can arise from the operation and interplay of the simple rules (building blocks) that compose a sampling strategy. Underscoring the merits of a cognitive-ecological perspective, the analyses also reveal how sampling strategies interact with the task environment, that is, how a given sampling strategy produces different choice behaviors depending on the properties of the choice ecology.