hostile media effect; (2) what factors about the source or message might erase it; (3) what
differences in information-processing mechanisms might underlie this bias?
Testing Reach and Source Factors
This paper focuses on the last of these, the information-processing question. However, to
do so requires that we also discuss the preceding questions in the context of an experimental test
of two rival theoretical explanations for the hostile media effect. A preliminary but crucial part of
the design for this study was the manipulation of two message characteristics – “reach” (the
number of people exposed to the message) and “source” (whether the author of the message was
a professional journalist or not). Results of previous studies suggested a possible role for one or
both factors (see e.g., Giner-Sorolla & Chaiken, 1994; Gunther & Schmitt, 2004; Arpan &
Raney, 2003).
We hypothesized that partisans would perceive a hostile bias in information presented in
a high-reach format such as a news article, but not in a low-reach format such as a student essay.
To test the source effect, we hypothesized that a hostile bias would be perceived when the author
of the information was presented as a journalist, but not as a college student. Analyses of these
characteristics have been reported elsewhere (authors, in press) but are described below in
abbreviated fashion because the processing mechanisms we tested are linked to these conditions.
The reach and source explanations for the hostile media effect address why the contrast bias
occurs, but interesting additional questions have been raised about how it happens. Thus, we also
wished to explore which information-processing mechanisms cause a partisan to see information
as disagreeable in one context, but not in another and which factor (reach or source) might
trigger any relevant mechanisms.
Testing Mechanisms
Convention