The Consequences of Validated and Self-Reported Voting Measures

In: Public opinion quarterly: journal of the American Association for Public Opinion Research, Band 45, Heft 4, S. 519-535

Abstract

Presented are research results comparing levels of error in survey Rs' self-reported registration & voting behavior, & independent verification of it. The consequences of these measurement errors for standard models of electoral participation & for estimates of the partisan division of the vote are assessed. Data come from 2 national surveys, each involving 2,300+ interviews with members of the US electorate, conducted in 1976 & 1978. These data were combined with field visits to local election administration offices to verify reported electoral participation. The analysis shows consistent misreporting of political behavior in both surveys of 12 percentage points for both registration status & voting. There was strong evidence of a bandwagon effect in the reported direction of the vote, favoring the winner but tempered by the candidate's relative electoral performance. No major changes were observed in the fundamental nature of basic relationships in 3 standard models of political participation when the dependent variable was switched from a self-reported to validated measure. 6 Tables, Appendix. AA.