Testing a Projection Explanation for Exemplar Effect

Kang Namkoong, Kwangjun Heo, Sojung Claire Kim, Timothy K F Fung, Albert Gunther

Research output: Contribution to conferenceConference paperpeer-review

Abstract

This study examines underlying mechanisms in the relationship between exemplification in mass media and people’s perception of public opinion. Previous research has consistently shown that exemplars influence the perception of current and future public opinion for controversial issues. Based on projection theory, we hypothesized that the effect of exemplification on estimates of public opinion change would be mediated by individuals’ own attitude toward an issue. To test the hypothesis, we employed the Sobel test and the Bootstrap method outlined by Preacher and Hayes (2004) as well as a series of Ordinary-Least Square (OLS) hierarchical regression analyses. Results show that the proportion of exemplars in a new article is positively associated with estimates of current and future public opinion. In addition, the relationship was fully mediated by the individuals’ attitude. In other words, vivid exemplars may have their strongest initial effect on the audience’s personal opinions. And personal opinion, in turn, strongly influences perceived public opinion via projection effects. This finding provides a new potential explanation for the process by which exemplification affects peoples’ estimations.
Original languageEnglish
Publication statusPublished - May 2009
Event59th Annual International Communication Association Conference, ICA 2009: Keywords in Communication - Chicago, United States
Duration: 21 May 200925 May 2009
https://convention2.allacademic.com/one/ica/ica09/ (Link to conference online programme)

Competition

Competition59th Annual International Communication Association Conference, ICA 2009
Country/TerritoryUnited States
CityChicago
Period21/05/0925/05/09
Internet address

Fingerprint

Dive into the research topics of 'Testing a Projection Explanation for Exemplar Effect'. Together they form a unique fingerprint.

Cite this