SOME LIKE IT HOT: INDIVIDUAL DIFFERENCES IN RESPONSES TO GROUP FEELING THERMOMETERS

In this paper we explore the nature and causes of individualdifferences in the use of feeling thermometer items to evaluatesocial groups. Some respondents have a general tendency to assignhigher overall temperature readings and some tend to use a widerrange of the temperature scale. These tendencies are partiallyunderstandable in terms of the respondent's evaluations of socialgroups, but other important predictors are identified. Methodsof adjusting feeling thermometer scores to account for individualdifferences are explored, using data from the 1984 AmericanNational Election Study. The conclusion is that some sort ofadjustment is probably useful for those studying supportersof conservative groups, though such a procedure may not be neededfor those who focus on liberal groups.