The politics of scale.

Subtle changes in the rating scales used by pollsters may markedly affect not only the extent to which people report liking a political candidate but also their ensuing estimates of how effective that person will be once elected.

Geoffrey Haddock and Rachael Carrick, both psychologists at the University of Exeter in England, asked 101 college students to rate British prime minister candidate Tony Blair on either of two scales for intelligence, honesty, and other personal characteristics the day before the 1997 general election. Regardless of prior political views, volunteers rated Blair much more favorably on a scale of -5 to +5 than on a scale of 0 to l0. Several days after Blair's election, students described in general terms their expectations. Those who had rated Blair on the -5 to +5 scale predicted that the new prime minister would do better than did students who had used the 0 to 10 scale, Haddock and Carrick report in the rail issue of SOCIAL COGNITION.

People may interpret negative numbers as describing a trait's opposite, such as the presence of dishonesty when asked about honesty. If so, on the -5 to +5 scale, any level of honesty gets rated from 0 to 5, yielding higher ratings than on a 0 to 10 scale, where numbers under 5 still imply honesty. This manipulation nudges other attitudes toward the politician in a positive direction, at least for a short while, the researchers assert.