Inference involves two settings: In the first, information is acquired (learning); in the second, it is applied (predictions or choices). Kind learning environments involve close matches between the informational elements in the two settings and are a necessary condition for accurate inferences. Wicked learning environments involve mismatches. This conceptual framework facilitates identifying sources of inferential errors and can be used, among other things, to suggest how to target corrective procedures. For example, structuring learning environments to be kind improves probabilistic judgments. Potentially, it could also enable economic agents to exhibit maximizing behavior.

Recently, researchers have investigated differences in decision making based on description and experience. We address the issue of when experience-based judgments of probability are more accurate than are those based on description. If description is well understood ("transparent") and experience is misleading ("wicked"), it is preferable to experience. However, if description is not transparent, will valid ("kind") experience lead to more accurate judgments? We report 2 experiments. The first involved 7 well-known probabilistic inference tasks. Participants differed in statistical sophistication and answered with and without experience obtained through sequentially simulated outcomes. The second experiment involved interpreting the outcomes of a regression analysis when making inferences for investment decisions. In both experiments, even the statistically naïve achieved accurate probabilistic inferences after experiencing sequentially simulated outcomes, and many preferred this presentation format. We conclude by discussing theoretical and practical implications.

No abstract

Providing information for decision making should be like telling a story. You need to know, first, what you want to say; second, whom you are addressing; and third, how to match the message and audience. However, data presentations frequently fail to follow these simple principles. To illustrate, we focus on presentations of probabilistic information that accompany forecasts. We emphasize that the providers of such information often fail to realize that their audiences lack the statistical intuitions necessary to understand the implications of probabilistic reasoning. We therefore characterize some of these failings prior to conceptualizing different ways of informing people about the uncertainties of forecasts. We discuss and compare three types of methods: description, simulation, and mixtures of description and simulation. We conclude by identifying gaps in our knowledge on how best to communicate probabilistic information for decision making and suggest directions for future research.SpanishMinisterio de Economía y Competitividadpublisher versio

No abstract