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 valuein election forecasting. However, statewide mail surveys conductedby the Columbus Dispatch newspaper since 1980 have made remarkablyaccurate forecasts of Ohio election outcomes. In comparisonto statewide surveys by two other organizations employing conventionaltelephone interview methods, the mail surveys were consistentlymore accurate and were generally less susceptible to sourcesof inaccuracy such as high rolloff and low publicity. The mailsurvey's advantage is attributable at least in part to largersample sizes, sampling and response procedures that yieldedmore representative samples of voters, lack of the need to allocateundecided respondents, and superior questionnaire design. Thesefindings suggest that mail surveys not only may be viable alternativesto telephone surveys but may actually be superior to them undersome conditions. Further-more, these results demonstrate thatsurveys with low response rates are not necessarily low in validity.     

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-opinionoption should reduce the pressure to give substantive responsesfelt by respondents who have no true opinions. By contrast,the survey satisficing perspective suggests that no-opinionoptions may discourage some respondents from doing the cognitivework necessary to report the true opinions they do have. Weaddress these arguments using data from nine experiments carriedout in three household surveys. Attraction to no-opinion optionswas found to be greatest among respondents lowest in cognitiveskills (as measured by educational attainment), among respondentsanswering secretly instead of orally, for questions asked laterin a survey, and among respondents who devoted little effortto the reporting process. The quality of attitude reports obtained(as measured by over-time consistency and responsiveness toa question manipulation) was not compromised by the omissionof no-opinion options. These results suggest that inclusionof no-opinion options in attitude measures may not enhance dataquality and instead may preclude measurement of some meaningfulopinions.     

AN EVALUATION OF A COGNITIVE THEORY OF RESPONSE-ORDER EFFECTS IN SURVEY MEASUREMENT

Previous research has documented effects of the order in whichresponse choices are offered to respondents using closed-endedsurvey items, but no theory of the psychological sources ofthese effects has yet been proposed. This paper offers sucha theory drawn from a variety of psychological research. Usingdata from a split-ballot experiment in the 1984 General SocialSurvey involving a variant of Kohn's parental values measure,we test some predictions made by the theory about what kindof response order effect would be expected (a primacy effect)and among which respondents it should be strongest (those lowin cognitive sophistication). These predictions are confirmed.We also test the "form-resistant correlation" hypothesis. Althoughcorrelations between items are altered by changes m responseorder, the presence and nature of the latent value dimensionunderlying these responses is essentially unaffected.     

EDUCATION MODERATES SOME RESPONSE EFFECTS IN ATTITUDE MEASUREMENT

On the basis of over 130 survey-based experiments, Schuman andPresser concluded that respondents' levels of edu cational attainmentdid not seem to be a pervasive and systematic moderator of susceptibilityto response effects. Via a meta analysis of these same data,we found that lower education was indeed associated with greaterstrength of seven response effects: response order effects,acquiescence, middle alternative effects not involving statusquo options, no-opinion filter effects, forbid/ allow effects,balance effects, and question order effects based on the normof reciprocity. The specific patterns of relations ob tainedare consistent with the notion that some of these response effectsmay result from satisficing.     

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-faceinterviewing with telephone interviewing as the dominant modeof survey data collection in the United States. But some ofthe most expensive and large-scale nationally funded, long-termsurvey research projects involving national area-probabilitysamples and long questionnaires retain face-to-face interviewingas their mode. In this article, we propose two ways in whichshifting such surveys to random digit dialing (RDD) telephoneinterviewing might affect the quality of data acquired, andwe test these hypotheses using data from three national modeexperiments. Random digit dialing telephone respondents weremore likely to satisfice (as evidenced by no-opinion responding,nondifferentiation, and acquiescence), to be less cooperativeand engaged in the interview, and were more likely to expressdissatisfaction with the length of the interview than were face-to-facerespondents, despite the fact that the telephone interviewswere completed more quickly than the face-to-face interviews.Telephone respondents were also more suspicious about the interviewprocess and more likely to present themselves in socially desirableways than were face-to-face respondents. These findings shedlight on the nature of the survey response process, on the costsand benefits associated with particular survey modes, and onthe nature of social interaction generally.     

The Measurement of Values in Surveys: A Comparison of Ratings and Rankings

Social values are most commonly measured using ranking techniques,but there is a scarcity of systematic comparisons between rankingsand other approaches to measuring values in survey research.On the basis of data from the 1980 General Social Survey, thisarticle evaluates the comparability of results obtained usingrankings and ratings of valued qualities. The comparison focuseson (1) the ordering of aggregate value preferences and (2) themeasurement of individual differences in latent value preferences.The two methods are judged to be similar with respect to orderingthe aggregate preferences of the sample, but dissimilar withregard to the latent variable structure underlying the measures     

A TEST OF THE FORM-RESISTANT CORRELATION HYPOTHESIS: RATINGS, RANKINGS, AND THE MEASUREMENT OF VALUES

A number of researchers have argued that ranking techniquesare more appropriate than rating methods for the measurementof values in surveys. The form-resistant correlation hypothesisproposes that observed associations among values and betweenvalues and other variables should remain invariant across measurementmethods. However, some recent research on parental values forchild qualities suggests that ratings and rankings produce differentcorrelational results. The present study tests the hypothesisthat discrepancies between rating and ranking results are dueto the fact that, when responding to rating questions, somerespondents avoid making difficult choices between valued qualitiesby rating all the qualities as highly and equally desirable.Consistent with this hypothesis, when nondifferentiating respondentsare removed from the analyzed sample, the substantive resultsof analyses of rating data resemble the results typically obtainedusing ranking data. This suggests that ranking may be the superiormethod for measuring values.