Coarse, Efficient Decision-Making
研究发现，决策者为降低决策成本应使用粗糙的分类标准，每个标准只设少量类别，即使需要更多标准，最有效的是使用二元标准，这可能是现实中广泛使用粗糙标准的原因，且二元标准能产生理性选择。
Abstract To minimize the cost of making decisions, an agent should use criteria to sort alternatives and each criterion should sort coarsely. To decide on a movie, for example, an agent could use one criterion that orders movies by genre categories, another by director categories, and so on, with a small number of categories in each case. The agent then needs to aggregate the criterion orderings, possibly by a weighted vote, to arrive at choices. As criteria become coarser (each criterion has fewer categories) decision-making costs fall, even though an agent must then use more criteria. The most efficient option is consequently to select the binary criteria with two categories each. This result holds even when the marginal cost of using additional categories diminishes to 0. The extensive use of coarse criteria in practice may therefore be a result of optimization rather than cognitive limitations. Binary criteria also generate choice functions that maximize rational preferences: decision-making efficiency implies rational choice.