The value of many important applications of survey data in academic research and public policy-making hinges on the accuracy of the assessments provided. This Small Grant for Exploratory Research (SGER) will collect data from the 1994 state-wide elections in Ohio as part of a field experiment to examine the impact of questionnaire format on the predictive value of mail surveys versus telephone interview methods for forecasting election results. Previous analyses of statewide elections in Ohio have demonstrated that self-administered mail surveys provide more accurate forecasts than telephone surveys. One possible explanation for this surprising result is that the mail surveys employed a different questionnaire format than the telephone surveys. This research will examine the impact of questionnaire format on accuracy by manipulating the form of the questionnaire administered by both mail and telephone modes immediately prior to the 1994 Ohio elections. In addition, this research will explore the impact of enhanced call-backs to telephone respondents on accuracy. This research effort should both advance basic survey measurement theory and provide a set of practical recommendations to analysts interested in maximizing measurement accuracy.
