Assessing the 1992 presidential and vice presidential debates: the public rationale.

• Author: Winkler, Carol K.
• Position: Special Issue: Political Campaign Debates

Since the 1960 presidential debate, the scholarly community, media, and public have devoted substantial energy to determining the winners and losers of presidential and vice presidential debates. This preoccupation has prompted criticism because it diverts attention from the substance available in public debates (Berquist & Golden, 1981; Jamieson & Birdsell, 1988). Notwithstanding the merits of such criticism, voters continue to assign victory and defeat in even the closest and most ambiguous debate contests (Sears & Chaffee, 1979).

The reasons voters use to explain their choices of winners and losers in presidential and vice presidential debates remain unclear. Numerous studies address whether personal attributes/presidential skills weigh more heavily than a discussion of the policy issues in the decision-making calculus of voters (see, for example, Berquist & Golden, 1981; Glass, 1985; Miller & MacKuen, 1979). The tendency to reduce the breadth of rationales employed by the electorate into a dichotomous framework (i.e., issue vs. image), however, prevents a more thorough understanding of the reasoning processes of the public.

Vancil and Pendell (1984) take a first step toward broadening the understanding of voter rationales in their analysis of the 1980 Carter-Reagan presidential debate. Basing their results on a survey of the presidential debate literature, a post hoc review of a limited telephone survey, and their own conceptual analysis, they propose six criteria for assessing performance in political debates. These include the pre-debate candidate preference, consistency with viewer positions on the issues, superior skills of advocacy, superior presidential personality, profiting from a major blunder of the opponent, and the media's choice of winner.

The methodology used in Vancil and Pendell's telephone survey, however, limited the public's responses to predetermined options spawned from previous research findings. For example, viewers were asked, "In viewing the debate between Presidential candidates, do you think it is most important to find out the candidate's stands on the issues, or to find out which candidate has the personality and leadership qualities a President should have?". While a complete accounting of the questions asked of respondents is not available, the weaknesses inherent in the study's methodology prompt even Vancil and Pendell to admit that the criteria they identify do not exhaust the possibilities.

By providing potential voters an open-ended opportunity to explain their reasons for choosing winners and losers in the first 1992 presidential debate and the 1992 vice presidential debate, this study expands understanding about the electorate's schemas. These debates constituted substantial campaign events as eighty-one million viewers watched the October 11 presidential debate and seventy-six million Americans observed the October 13 vice presidential debate (Carmody, 1992).

In addition to discovering the reasons given by the public for choosing winners and losers, this study seeks to determine if various subgroups of the electorate use different reasons for judging debate performance. Previous research has indicated that educational status (Glass, 1985), party affiliation (Campbell, et al., 1960), audience predisposition (Leuthold & Valentine, 1982; Sears & Chaffee, 1979; Sigelman & Sigelman, 1984), and media coverage (Lang & Lang, 1978; Lanoue & Schrott, 1991; Lowery, Bridges, & Barefield, 1990; Miller & MacKuen, 1979) influence voter perceptions of candidate performance, but none has specifically addressed whether the rationales for choosing winners and losers varies according to these factors. This study analyzes whether viewership of the debates, demographic groupings, party affiliation, media usage, and choice of particular candidates as winners or losers constitute subgroups that use different reasoning in assessing candidate performance.

The rationales used by various sub-groups of the public to choose debate winners and losers not only have practical applications for those who would participate in the political arena, but can also provide a substantive base for critically evaluating shifts in electoral thought. An indepth understanding of the reasons voters choose particular winners and losers in political debates could yield insight into sensible debate format changes, alterations in media reporting, and variations in the content of the questions that panelists pose in such debates.

METHODOLOGY

From October 14 to October 28, 1992, a survey was administered to 27 undergraduate and graduate Communication classes at Georgia State University. Located in downtown Atlanta, Georgia State is a commuter university which services both inner-city and suburban residents.

The survey asked respondents to indicate demographic information (age, sex, and race), party affiliation, viewership of the October 11 presidential debate and/or the October 13 vice presidential debate, and any sources of media they used that covered the debates (newspapers, television networks, radio, and magazines). The survey then asked each respondent to indicate who, in their opinion, had won or lost the presidential and vice presidential debates. Respondents had the option to report that no one had won or lost the debates or to choose more than one candidate as the winner(s) or loser(s) of the debates. In an effort to avoid biasing the subjects' reasoning process, an open-ended question asked participants to explain their reason(s) for their choices of winner(s) and loser(s). Three hundred and seventy respondents provided demographic, party affiliation, viewership, sources of media coverage and choice of winner(s)/loser(s) data; of these, 33 failed to provide reasons for their assessments of the candidates' performance.

Recurrent responses were grouped into thirteen categories for coding purposes. These included confidence/presence, specificity of response to posed questions, honesty/trustworthiness, use of an attack strategy, level of preparation, connectedness to people's issues, proposal of new solutions, advocacy of specific policies or political party, leadership ability, ability to exceed expectations, level of participation, ability to appear presidential, and intelligence. Each category included statements that indicated a presence or absence of the characterization (e.g., responses that indicated that a candidate attacked too much, too little, or effectively were all coded as instances of the "use of the attack strategy" category).

Inter-coder reliability was established by using a graduate student uninvolved with the project to code a random sample of 10% (n = 37) of the surveys. Coding instructions were the following: "confidence/presence" included whether the candidates were in command and whether their delivery/public speaking skills were perceived as effective; "specificity of response" included whether the candidates' answers directly addressed the questions posed by the panelists; "honesty/trustworthiness" encompassed the perception of the candidates' sincerity, consistency of position, and level of manipulation; "use of an attack strategy" included comments regarding the aggressiveness or defensiveness of the candidates; "connectedness to people's issues" encompassed whether the candidates demonstrated concern for the public's problems, and "level of preparation" included discussions of whether the candidates adequately prepared for the debate.

"Proposal of new solutions" encompassed praise or blame for the candidates' general ability to solve the nation's problems. "Advocacy of a specific policy/party," by contrast, included commentary regarding single issues or alignment with a particular candidate's party affiliation. "Leadership" encompassed whether the candidates possessed...