Knowing When to Stop and Make a Choice, an Experiment on Optimal Sequential Sampling

Abstract

We investigate how people make choices when they are unsure about the value of the options they face, and can invest in the acquisition of more information. We design a laboratory experiment to study whether human behaviour is able to approximate the optimal solution to this sequential sampling problem. We find that participants deviate from the optimal strategy in a systematic manner: information is either mostly over-sampled or mostly under-sampled, depending on the cost of information. With time, participants tend to learn, without converging towards equilibrium. This deviation costs participants between 10% and 25% of their potential payoffs. We examine various explanations to our results, and find that the most likely one is that participants exhibit a confirmation bias.

Keywords: Search, Decision, Applications, Information, Economics

JEL Classification: C91, D81, D83