Beyond the Notion of Accessibility Bias: Message Content as the Common Source of Agenda-Setting and Priming Effects

ABSTRACT

This study examines the cognitive mechanisms behind agenda-setting and priming effects. Recent evidence suggessts that accessibility effects within network models of memory are not well suited to explain agenda-setting and priming effects. This article attempts to provide additional evidence regarding the roles of issue accessibility and message content in agenda-setting and priming processes. Our findings indicate that changes in issue accessibility are not a sufficient condition for agenda-setting and priming effects. Instead, the content of mediated messages that validates an issue as a matter of public concern is shown to be the primary explanation of agenda-setting and priming effects. Different psychological processes observed in social psychological studies and political communication studies as well as theoretical relationships between agenda setting and priming are discussed.

Notes

¹ The experiment was conducted before the Pyeongchang Winter Olympics where tensions began to lessen (“North Korea Profile”, 2019).

² Measuring accessibility and ASP outcomes one after the other may contaminate answers to the following measures because these measures may inflate accessibility (Miller, 2007).

³ There was no association between whether or not respondents cited crime as the MIP and the presentation order, χ² (1) = .01, p = .91. Additionally, the presentation order did not make a significant difference in thermometer ratings (M = 77.42, SE = 2.19 for agenda setting first; M = 78.26, SE = 2.34 for priming first), t(65) = − .54, p = .59, or evaluations of the president’s handling of the issue (M = 4.89, SE = .19 for agenda setting first; M = 4.48, SE = .26 for priming first), t(65) = 1.28, p = .21.

⁴ The experiment was approved by the Institutional Review Board at the University of Wisconsin-Madison on October 25, 2016.

⁵ Profile likelihood confidence intervals are reported for general linear models (Hosmer, Lemeshow, & Sturdivant, 2013). Thus, the values in these confidence intervals represent odds ratios. All others are conventional Wald confidence intervals.

⁶ The experiment was approved by the Institutional Review Board at the University of Wisconsin-Madison on November 21, 2017.

⁷ The association between whether or not respondents cited North Korea crisis as the MIP and the presentation order was not significance, χ² (1) = 1.71, p = .19, indicating that responses to MIP questions were not contaminated by increased issue accessibility, resulting from answering the issue approval question. Also, the presentation order did not make a significant difference in presidential approval rating (M = 28.16, SE = 3.41 for agenda setting first; M = 21.24, SE = 2.82 for priming first), t(147) = 1.56, p = .12, and in evaluations of the president’s handling of the issue (M = 2.69, SE = .19 for agenda setting first; M = 2.44, SE = .17 for priming first), t(155) = .99, p = .32.

Additional information

Notes on contributors

ByungGu Lee

ByungGu Lee has a Ph.D. in mass communication from the University of Wisconsin-Madison.

Jiawei Liu

Jiawei Liu is a Postdoctoral Associate in the Department of Communication at Cornell University.

Hyesun Choung

Hyesun Choung is a doctoral student in the School of Journalism and Mass Communication at the University of Wisconsin-Madison.

Douglas McLeod

Douglas McLeod is Evjue Centennial professor at the University of Wisconsin-Madison.