HOW TO MEASURE GENERALIZED TRUST: WORDINGS, SCALES, AND DON’T KNOWS

• Recently, survey institutes have changed their measurement of generalized trust from the standard dichotomy to an eleven-point scale. Additionally, these survey institutes use different wordings and some of them provide a “don’t know” option. Using two survey based experiments, one on 12,000 self-selected Swedes and the other on a probability-based sample of 3,000 Swedes, this article evaluates the generalized trust question in terms of wordings, scales, and inclusion of a don’t know option. The findings suggest that, in line with the advice from Krosnick and Fabrigar (1997), generalized trust is as good measured using a minimally balanced question wording as a fully balanced; is best measured with a seven- or eleven-point scale with a neutral midpoint; and is best measured without including a don’t know option.

Ämnesord

SAMHÄLLSVETENSKAP  -- Statsvetenskap -- Statsvetenskap (hsv//swe)

SOCIAL SCIENCES  -- Political Science -- Political Science (hsv//eng)

Nyckelord

Generalized trust

Question Wording

Scales

Don’t knows

Survey methodology

Publikations- och innehållstyp

vet (ämneskategori)

kon (ämneskategori)