The way in which information about proportions, amounts, frequencies, probabilities, degrees of con®dence, and risk is portrayed in natural language is not neutral, but re¯ects pre- suppositions and assumed norms. In this paper we present a review of evidence in support of this position. We show that the choice of expressions for communication depends in a systematic way on the kinds of inferences communicators draw. We go on to discuss the consequences of this for attribution phenomena, aspects of reasoning, the portrayal of uncertainty, and responses to questionnaires. We also suggest that communicator preferences for using language rather than numbers may have to do with human reasoning being argument-based, rather than with a preference for vagueness, as has been commonly claimed. Copyright # 2000 John Wiley & Sons, Ltd.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below