Jon A. Krosnick

Attitude strength : antecedents and consequences

Contents: P.E. Converse, Foreword. Preface. J.A. Krosnick, R.E. Petty, Attitude Strength: An Overview. R.P. Abelson, Attitude Extremity. C.M. Judd, M. Brauer, Repetition and Evaluative Extremity. A. Tesser, L. Martin, M. Mendolia, The Impact of Thought on Attitude Extremity and Attitude-Behavior Consistency. R.E. Petty, C.P. Haugtvedt, S.M. Smith, Elaboration as a Determinant of Attitude Strength: Creating Attitudes That Are Persistent, Resistant, and Predictive of Behavior. W.D. Crano, Attitude Strength and Vested Interest. D.S. Boninger, J.A. Krosnick, M.K. Berent, L.R. Fabrigar, The Causes and Consequences of Attitude Importance. C.J. Thomsen, E. Borgida, H. Lavine, The Causes and Consequences of Personal Involvement. S.R. Gross, R. Holtz, N. Miller, Attitude Certainty. R.H. Fazio, Attitudes as Object-Evaluation Associations: Determinants, Consequences, and Correlates of Attitude Accessibility. W. Wood, N. Rhodes, M. Biek, Working Knowledge and Attitude Strength: An Information-Processing Analysis. A.R. Davidson, From Attitudes to Actions to Attitude Change: The Effects of Amount and Accuracy of Information. J. Jaccard, C. Radecki, T. Wilson, P. Dittus, Methods for Identifying Consequential Beliefs: Implications for Understanding Attitude Strength. M.M. Thompson, M.P. Zanna, D.W. Griffin, Lets Not Be S. Chaiken, E.M. Pomerantz, R. Giner-Sorolla, Structural Consistency and Attitude Strength. A.H. Eagly, S. Chaiken, Attitude Strength, Attitude Structure, and Resistance to Change. M.W. Erber, S.D. Hodges, T.D. Wilson, Attitude Strength, Attitude Stability, and the Effects of Analyzing Reasons. D.T. Wegener, J. Downing, J.A. Krosnick, R.E. Petty, Measures and Manipulations of Strength-Related Properties of Attitudes: Current Practice and Future Directions.

Telephone versus Face-to-Face Interviewing of National Probability Samples with Long Questionnaires: Comparisons of Respondent Satisficing and Social Desirability Response Bias

The last 50 years have seen a gradual replacement of face-to-face interviewing with telephone interviewing as the dominant mode of survey data collection in the United States. But some of the most expensive and large-scale nationally funded, long-term survey re- search projects involving national area-probability samples and long questionnaires retain face-to-face interviewing as their mode. In this article, we propose two ways in which shifting such surveys to random digit dialing (RDD) telephone interviewing might affect the quality of data acquired, and we test these hypotheses using data from three na- tional mode experiments. Random digit dialing telephone respondents were more likely to satisfice (as evidenced by no-opinion responding, nondifferentiation, and acquiescence), to be less cooperative and en- gaged in the interview, and were more likely to express dissatisfaction with the length of the interview than were face-to-face respondents, despite the fact that the telephone interviews were completed more quickly than the face-to-face interviews. Telephone respondents were also more suspicious about the interview process and more likely to present themselves in socially desirable ways than were face-to-face respondents. These findings shed light on the nature of the survey re- sponse process, on the costs and benefits associated with particular survey modes, and on the nature of social interaction generally.

Attitude strength: One construct or many related constructs?

A variety of attributes differentiate attitudes that are stable and conseguential from those that are not, including extremity, certainty, importance, knowledge, intensity, interest, direct experience, accessibility, latitudes of rejection and noncommitment, and affective-cognitive consistency. Although these dimensions are clearly conceptually and operationally distinct from one another, researchers have often assumed that some are interchangeable, or that two or more reflect common higher-order constructs. Three studies using multitrait-multimethod confirmatory factor analysis assessed the relations among these dimensions. Although some of these dimensions are strongly related, most are not, and a multifactor model seems necessary to account for their intercorrelations

Aging and Susceptibility to Attitude Change

Two hypotheses about the relation between age and susceptibility to attitude change were tested. The impressionable years hypothesis proposes that individuals are highly susceptible to attitude change during late adolescence and early adulthood and that susceptibility drops precipitously immediately thereafter and remains low throughout the rest of the life cycle. The increasing persistence hypothesis proposes that people become gradually more resistant to change throughout their lives. Structural equation models were applied to data from the 1956-1960, 1972-1976, and 1980 National Election Panel Studies in order to estimate the stability of political attitudes and unreliability in measures of them. The results support the impressionable years hypothesis and disconfirm the increasing persistence hypothesis. A decrease in the over-time consistency of attitude reports among 66- to 83-year-olds was found to be due to increased random measurement error in their reports, not to increased attitude change.

Government policy and citizen passion: A study of issue publics in contemporary America

This article describes the findings of a program of research exploring the cognitive and behavioral consequences of passionate concern about government policy issues. American citizens vary a great deal in terms of the personal importance they attach to their attitudes on particular policy issues. Citizens whose policy attitudes are especially important to them are likely to think frequently about those attitudes, to perceive competing candidates as being relatively polarized on the issue, and to form presidential candidate preferences on the basis of those attitudes. Also, policy attitudes that citizens consider personally important are highly resistant to change and are therefore especially stable over long periods of time. The American public appears to be structured into many small issue publics, each composed of citizens who are passionately concerned about a single issue. Most Americans fall into very few issue publics, the particular ones being determined by each individuals unique self-interests, social identifications, and cherished values. The implications of these findings for the workings of democracies are discussed.

Origins of attitude importance, self-interest, social identification, and value relevance

Five studies examined the relations between attitude importance and 3 of its hypothesized determinants: self-interest, social identification with reference groups or reference individuals, and cherished values. Verbal protocols, multivariate analysis of survey data, and laboratory experimentation revealed that (1) peoples theories of the causes of attitude importance pointed to all 3 hypothesized predictors, (2) the 3 predictors each had significant, unique statistical associations with importance, and (3) a manipulation of self-interest yielded a corresponding change in importance. These results help clarify the nature and origins of attitude importance, challenge the widely believed claim that self-interest has little or no impact on political cognition, and identify new likely consequences of social identification processes and values.

The Impact of "No Opinion" Response Options on Data Quality: Non-Attitude Reduction or an Invitation to Satisfice?

According to many seasoned survey researchers, offering a no-opinion option should reduce the pressure to give substantive re- sponses felt by respondents who have no true opinions. By contrast, the survey satisficing perspective suggests that no-opinion options may dis- courage some respondents from doing the cognitive work necessary to report the true opinions they do have. We address these arguments using data from nine experiments carried out in three household surveys. Attraction to no-opinion options was found to be greatest among re- spondents lowest in cognitive skills (as measured by educational at- tainment), among respondents answering secretly instead of orally, for questions asked later in a survey, and among respondents who devoted little effort to the reporting process. The quality of attitude reports ob- tained (as measured by over-time consistency and responsiveness to a question manipulation) was not compromised by the omission of no- opinion options. These results suggest that inclusion of no-opinion op- tions in attitude measures may not enhance data quality and instead may preclude measurement of some meaningful opinions.

The association of knowledge with concern about global warming: trusted information sources shape public thinking.

During the last decade, a great deal of news media attention has focused on informing the American public about scientific findings on global warming (GW). Has learning this sort of information led the American public to become more concerned about GW? Using data from two surveys of nationally representative samples of American adults, this article shows that the relation between self-reported knowledge and concern about GW is more complex than what previous research has suggested. Among people who trust scientists to provide reliable information about the environment and among Democrats and Independents, increased knowledge has been associated with increased concern. But among people who are skeptical about scientists and among Republicans more knowledge was generally not associated with greater concern. The association of knowledge with concern among Democrats and Independents who trust scientists was mediated by perceptions of consensus among scientists about GWs existence and by perceptions that humans are a principal cause of GW. Moreover, additional analyses of panel survey data produced findings consistent with the notion that more knowledge yields more concern among Democrats and Independents, but not among Republicans. Thus, when studying the relation of knowledge and concern, it is important to take into account the content of the information that different types of people acquire and choose to rely upon.

MAIL SURVEYS FOR ELECTION FORECASTING? AN EVALUATION OF THE COLUMBUS DISPATCH POLL

Because of slow turnaround time and typically low response rates, mail surveys have generally been considered of little value in election forecasting. However, statewide mail surveys conducted by the Columbus Dispatch newspaper since 1980 have made remarkably accurate forecasts of Ohio election outcomes. In comparison to statewide surveys by two other organizations employing conventional telephone interview methods, the mail surveys were consistently more accurate and were generally less susceptible to sources of inaccuracy such as high rolloff and low publicity. The mail surveys advantage is attributable at least in part to larger sample sizes, sampling and response procedures that yielded more representative samples of voters, lack of the need to allocate undecided respondents, and superior questionnaire design. These findings suggest that mail surveys not only may be viable alternatives to telephone surveys but may actually be superior to them under some conditions. FurtherPENNY S. VISSER is a doctoral candidate in the department of psychology at Ohio State University and JON A. KROSNICK is an associate professor in the departments of psychology and political science at Ohio State University. JESSE MARQUETTE iS a professor of political science and director of the Survey Research Center at the University of Akron. MICHAEL CURTIN iS the editor of the Columbus Dispatch. This research was supported by grant SBR-9503822 from the National Science Foundation to Jon A. Krosnick and by grant T32-MH19728-03 from the National Institute of Mental Health, which provided a predoctoral fellowship to Penny S. Visser. We thank Andy Smith and Al Tuchfarber from the University of Cincinnati for providing information on their preelection surveys, and Jim Hunter from the Columbus Dispatch for archival assistance. We also thank Richard Carson, Robert Cameron Mitchell, and Stanley Presser for helpful comments on an earlier draft. Correspondence regarding this article should be addressed to Penny S. Visser or Jon A. Krosnick, Department of Psychology, Ohio State University, 1885 Neil Avenue, Columbus, OH 43210 (E-mail visser.104@osu.edu or krosnick@osu.edu). Public Opinion Quarterly Volume 60:181-227 ? 1996 by the American Association for Public Opinion Research All rights reserved. 0033-362X/96/6002-0015

Attitude importance and attitude change

02.50 This content downloaded from 157.55.39.174 on Wed, 31 May 2017 18:33:52 UTC All use subject to http://about.jstor.org/terms 182 Visser, Krosnick, Marquette, and Curtin more, these results demonstrate that surveys with low response rates are not necessarily low in validity. Since their inauspicious beginnings over 150 years ago (Smith 1990), preelection forecasting surveys have undergone tremendous refinement (Field 1983; Frankovic 1992; Gallup 1972; Mann and Orren 1992; Perry 1960, 1979; Rosenstone 1983). Early surveys, such as the straw polls of the Literary Digest magazine, were characterized by haphazard sampling methods that precluded consistently accurate forecasts (Converse 1987; Field 1983; Gallup 1972; Rosenstone 1983; Squire 1988). But by the mid1930s, Gallup, Roper, and Crossley had begun using more scientific sampling techniques to ensure that their respondents accurately represented the electorate. These and other improvements allowed Gallup to forecast the U.S. presidential elections from 1936 to 1956 with an average error of 3.9% (Rosenstone 1983). Further methodological changes in the 1950s and 1960s (Perry 1960, 1979) brought the Gallup forecasts of the national elections from 1960 through 1980 to within an average of 1.6% of the actual election results (Rosenstone 1983). Perhaps the most sweeping change made by Gallup and other preelection survey outfits has been the shift from face-to-face interviewing to telephone interviewing. Of the 430 preelection surveys examined in one recent study, fully 98% interviewed by telephone (Crespi 1988). Telephone interviews are less expensive to conduct than face-to-face interviews, and the former typically achieve higher response rates than most mail surveys, leaving them less vulnerable to nonresponse bias. Furthermore, telephone interviewing can be completed much more quickly than face-to-face and mail surveys, which permits up-to-theminute tracking of candidate support. Despite these presumed improvements in procedures, however, preelection surveys still predict some election results with significant error. In 1980, for example, all four of the major polls underestimated Ronald Reagans margin of victory over Jimmy Carter, in one case by as much as 9% (Kagay 1992). Similarly, in the 1970 British election, four of the five final polls and 19 out of 20 earlier polls placed the Labor Party in the lead (by an average of 4.3%), whereas the Conservative Party won by 2.4% (Crewe 1992). More recently, last-minute forecasts of the 1993 New Jersey governors race showed incumbent Jim Florio leading challenger Christine Todd Whitman by as much as 15%, but on Election Day Florio was unseated (Gray 1993). Of course, even in a perfectly executed survey, one would expect some discrepancy between the election forecast and the actual outcome due to sampling error, but the systematic discrepancies cited above cannot be accounted for by sampling error alone. Clearly, there is room for further improvement of preelection survey methodology. This content downloaded from 157.55.39.174 on Wed, 31 May 2017 18:33:52 UTC All use subject to http://about.jstor.org/terms Mail Surveys for Election Forecasting? 183 Less clear, though, is how such improvements can best be achieved. In this article, we propose the consideration of self-administered mail surveys for use in election forecasting. We suggest that mail survey methodology, along with scientific sampling techniques and careful questionnaire design, may provide uniquely desirable conditions for accurately forecasting election outcomes. We begin by detailing the methods of preelection mail surveys conducted by the Columbus Dispatch newspaper in Ohio since 1980 and preelection telephone surveys by the University of Cincinnati and the University of Akron. We then compare the accuracy of the polls in forecasting the national and statewide elections every even year from 1980 to 1994 and explore the correlates of forecast accuracy. Finally, we test several possible explanations for the observed differences between the two survey modes.