Key points are not available for this paper at this time.
A news report on the plight of a minority of American family farmers was manipulated to create versions differing in the degree of precision of general information (precise, imprecise) and in the use of exemplifying case histories (selective, blended, representative). Selective exemplification featured only histories of failing farms, representative exemplification a distribution of histories of failing and successful farms proportional to their actual occurrence. Respondents reported their own views concerning the farmers' plight either shortly after reading or after a delay of one or two weeks. The accuracy of estimates of failing farms was found to be highest for representative and lowest for selective exemplification—despite the availability of corrective general information. These effects were stable over time.
Zillmann et al. (Sat,) studied this question.