What We Now Know about "I Don't Knows"

In: Public opinion quarterly: journal of the American Association for Public Opinion Research, Volume 39, Issue 2, p. 207-218

Abstract

Most survey research involves a large number of schedules with nonresponses, but no commonly agreed upon guidelines exist for handling the phenomenon in analysis. Part of the reason for this lack of guidelines may lie in the very conception of the phenomenon as random. Evidence presented here suggests that nonresponse is not a random phenomenon but is systematically related to definite R characteristics. An index of nonsubstantive response was developed for R's interviewed in connection with the 1960, 1964, & 1968 national election studies (N=1,181, 1,834, & 1,673 respectively) conducted by the U of Michigan's Survey Research Center. The index was found to be systematically related to R's age, education, income, sex, race, political involvement, & an index of political efficacy. In general it is hypothesized that nonsubstantive responses like "don't know" are a response style evinced by people excluded from information, influence, & decision-making processes. 7 Tables. AA.