SCALING INDIVIDUALS: QUESTIONNAIRE
DESIGN AND RATING SCALE CONSTRUCTION
Studying people's beliefs, attitudes, values, and personalities is a central research preoccupation of the social sciences. This focus on differences among individuals requires that data be organized and conceptualized in a systematic and precise manner, a manner different from that used in stimulus scaling, which we discussed in chapter 14. Whereas we assume in the methods of pair comparison and rank order scaling that all participants would respond identically to the various choice stimuli were it not for random error, the scaling of individual differences requires that we regard variations in responses among participants as meaningful (i.e., as not attributable to error). Conversely, the stimuli used to assess individual differences (typically the questions or “items” in the scale or questionnaire) are assumed to be identical in meaning for all participants, and to be measuring essentially the same idea, knowledge base, or attitude. Differences in participants' responses to these hypothetically “identical” items are the central methodological focus in the scaling of individuals.
Two complementary approaches characterize attempts to establish measures that assess differences among people. For convenience, we will term the slightly less formal measures questionnaires, and the more rigorously designed measures rating scales. The design tactics used in each help inform the other. Questionnaire and scale construction are integral to many of the data collection methods we discuss throughout this text, so it is important to have a good understanding of their strengths and weaknesses.
In questionnaires, we often do not have the luxury of length. That is, we are unable to use many different items to tap a person's evaluations of a given target such as a person, event, or object. Sometimes, the limitation is imposed because of cost. When conducting a national survey, for example, including even one extra item can be prohibitively expensive, so we must be content with an item or two to assess people's thoughts and feelings on a