What if the expected is not the most likely outcome? Four examples giving pause for thought and reconsideration
探讨了当分布不对称时，使用预期结果而非最可能结果作为分析焦点可能带来的偏差，通过四个建模案例说明视角选择对结论的实质性影响。
Abstract The foundational nature of expectations‐based theories and the prominence of symmetric unimodal stochastic assumptions in economic research render the expected outcome the go to locational focus throughout its many realms. When symmetric unimodality prevails, expected and most likely outcomes are identical; however, when it does not, they are not, begging the question ‘Why not use most likely rather than expected outcomes as the locational focus if there is potential for asymmetry?’ Here, the issue is explored by examining what drives the selection of locational foci by characterizing the choice as the solution to an analyst's optimization problem that reflects their perspective and typology. The choices are then discussed in the context of various economic modelling scenarios. The issue is generic to a wide variety of analytic scenarios, and its relevance hinges on the prevalence and extent of distributional asymmetries in those scenarios. To emphasize its diversity, four disparate modelling contexts where locational focus matters are examined. All demonstrate that possessing a ‘most likely’ as opposed to an ‘expectations‐based’ perspective makes a substantive difference to the analysis and conclusions drawn, which should give pause for thought.