Going Against the Grain: The Intersection of Framing, Source Credibility, and Predispositions

Download data and study materials from OSF

Principal investigator:

Kevin Arceneaux

Temple University

Email: kevin.arceneaux@temple.edu

Homepage: http://astro.temple.edu/~arceneau/

Sample size: 1268

Field period: 1/6/2005-1/12/2005


Political scientists have often mollified concerns about the ability of elites to manipulate public opinion by noting that citizens can align with factions that share their policy preferences (Schattschneider 1960; Zaller 1992). Research on framing effects confirms that individuals resist attempts to shape their opinions via frames when the message is delivered by a source from an opposing faction. Yet if citizens do rely on source cues when forming opinions and trusted elites take policy positions that are at variance with citizens' underlying predispositions, the conditions are ripe for elite manipulation to occur. Using experimental data collected from a nationally representative sample of adults, this paper explores whether subjects rely on their underlying predispositions or source cues when evaluating political messages and political messengers. The findings show that subjects resist attempts to frame issues in ways that are at odds with their underlying predispositions when forming opinions about the issue, but not necessarily when forming opinions about the messenger. Reliance on source cues appears to attenuate the extent to which underlying predispositions influence candidate evaluations, especially when the subject lacks political sophistication or the issue is technical in nature. The implications of these findings are discussed.


1) Predispositions will neutralize message framing by trust sources when the issue content of the frame is easy to grasp.
2) Source cues will dominate predispositions when the issue content of the frame is difficult to grasp.
3) Individuals with low levels of political sophistication will rely more on source cues when forming an opinion.
4) Individuals with higher levels of political sophistication will rely more on their predispositions when forming an opinion.

Experimental Manipulations

Subjects read a mock news story about a campaign speech given by a congressional candidate. The following elements of the article were manipulated:
1) Topic of the speech (abortion/federal responsibility for environmental regulation issue)
2) Party affiliation of the candidate (Democrat/Republican)
3) Candidate's stance (pro/anti)


1) Issue Opinion
2) Support for the candidate
3) Evaluation of how well the candidate represents the views of the respondent