The devil is in the details: Incorrect intuitions in optimal search

In the classic Secretary Problem it has been established that people tend to search somewhat less than is optimal, and a number of explanations have been suggested. Here we propose a new explanation, the Similar-But-Incorrect Intuitions Hypothesis, which says that suboptimal search behavior is to be expected because optimal strategies vary disproportionately with subtle details of the search problem setup, whereas people seem to entertain general intuitions about optimal search. We find support for this hypothesis in experiments on a new search problem, the Explore-and-Collect Problem, where the player collects utility from an option every time it is tried and options can be recalled. Although the optimal search effort in this problem is much smaller than for the Secretary Problem, people tend to search only marginally less. This is not predicted by previous explanations for suboptimal search.