Exaggerated Versus Representative Exemplification in News Reports

Abstract

A news report on carjacking, presented in magazine format, was manipulated to create versions differing in exemplar distortion (minimally, mildly, substantially, and extremely) and precision of base rate information (imprecise, precise). Time of assessment (immediate, delayed) was varied as well. Readers evaluated carjacking as a national issue, as likely to worsen, as a local danger, and as a personal threat. They also estimated the proportion of carjacking victims suffering injury. The readers eventually rated aspects of the news report. Readers of news featuring extreme exemplar distortion considered carjacking to be a more serious national problem than did readers of news featuring minimally, mildly, or substantially distorted exemplars. This effect was uniform for precision of base rate information and was stable over time. Readers presented with exemplars of people killed during carjackings grossly overestimated the incidence of such an outcome. Additionally, they found the report more upsetting than did readers presented with less extreme exemplars. There were no significant differences, however, in evaluations of newsworthiness, importance, or accuracy of the various versions of the report.