Question-Order Effects of Presidential Approval Ratings on Gubernatorial Approval Ratings: A Research Note

Abstract

Data from four Ohio polls, conducted between June 1983 and May 1984, showed a significant question-order effect of presidential approval on gubernatorial approval. The Governor of Ohio, Richard Celeste, received significantly higher approval ratings when the gubernatorial approval question was asked just before the presidential approval question than when it was asked immediately afterward. The analysis also suggests that the order effect on gubernatorial ratings may be due, in large part, to a tendency for strong Reagan partisans to contrast Celeste's job performance to that of Reagan's when the question about Reagan's job performance was askedfirst. The results have significant implications for the ordering of questions on sociopolitical objects or figures within a survey interview. Public opinion researchers have demonstrated repeatedly that the order and context in which a question is asked can significantly affect the results of a survey (see e.g., Bishop, Oldendick & Tuchfarber 1982, 1984a, 1984b; McFarland 1981; Schuman & Presser 1981). The purpose of this article is to examine question-order effects on presidential and gubernatorial approval (cf. Crespi & Morris 1984; Sigelman 1981) and to suggest a theoretical scheme for understanding how and why such effects occur. We will also discuss the implication of these findings for the design of survey questionnaires.