Agenda Setting for the Civil Right Issue

Abstract

A comparison of front-page New Yor-k Times'content and national public opinion from 1954 to 1976 showed strong agenda-setting effects for the civil rights issue. For this issue, the optimal effect span was the fourto six-week period immediately prior to field work. These findings contradict previous findings and assertions about a cumulative media effect over a longer period of time. James P. Winter is an Adjunct Lecturer in the Newhouse School of Communications, Syracuse University. Chaim H. Eyal is on the faculty of the Communications Institute at Hebrew University, Jerusalem. Public Opinion Quarterly Vol. 45:376-383 ? 1981 by The Trustees of Columbia University Published by Elsevier North-Holland, Inc. 0033-362X/81/0045-376/$2.50 This content downloaded from 157.55.39.144 on Wed, 07 Sep 2016 05:32:21 UTC All use subject to http://about.jstor.org/terms AGENDA SETTING FOR THE CIVIL RIGHTS ISSUE 377 case in the literature. The question of the appropriate for analysis has received little systematic attention; in fact it has only recently been elaborated (Eyal et al., 1981; Eyal, 1980). One of the time frame components identified is the optimal effect span or peak association between media and public emphasis of an issue. Those studies that have to some degree broached the questions of time frame suggest that the optimal effect span between aggregate media attention and public priority is between two and five months, and that the impact is a cumulative one, with exposure over time leading to enhanced public salience (McCombs, et al., 1975). As McCombs and Masel-Walters (1976) indicated: It appears that the cumulative effects of from three to four months of day-to-day news play result in some issues rising high on the agenda and others disappearing from public view (p. 7). In a nonelection study of university students, Stone (1975) examined Time and Newsweek for six months before and three months after the dates of his fieldwork. Accumulating the media duration backward from the interviews, he found a monotonic increase in correlations between media and public agendas, especially up to two months prior to interviewing. Cumulative media content from a full seven months prior to the interviews provided the highest zero-order correlations with students' agendas, on several issues. Despite this indication, some researchers have found agenda-setting effects using as little as one week's media content from immediately prior to the interview period (Mullins, 1977; Becker and McCombs, 1977). Indeed, in a study comparing national television and Gallup Poll data over an eight-year period, Zucker (1978) found that the media emphasis in the month immediately prior to the interview period was a better predictor of public opinion than was earlier media content. However, Zucker's study was designed to focus on the causal order in agenda setting, and on types of issues and their duration of exposure in the media rather than the optimal effect span. A second important variable that has only recently been considered is the nature of the individual issues examined. McCombs (1981) described four approaches to agenda-setting research: using either aggregate or individual public agenda data, in conjunction with either a set of issues or a single issue. But perhaps the variable nature of issues precludes treating them in the aggregate, another problem that may explain inconsistent findings. With a few recent exceptions (Zucker, 1978; Stroman, 1978; Erbring et al., 1980; Winter et al., 1980) agenda-setting researchers have aggregated diverse issues, and expected wholesale transferral of issue saliences from media to public. The problems associated with treating issues in the aggregate have This content downloaded from 157.55.39.144 on Wed, 07 Sep 2016 05:32:21 UTC All use subject to http://about.jstor.org/terms