Direction of Response Sale Format in International Comparison
Institute for Sociology University of Giessen Ludwigshafen, Germany
Sociology of Health Ben Gurion University of the Negev Beer Sheva, Israel
Direction of response scale format and response behavior
Formatting an answer - as one step in the cognitive process of answering a question - means to map the selected answer to the categories on a response scale. In this process response scales, as one part of questions, serve as orientation for respondents in mapping their answers to response categories. Therefore it can be expected that the direction of a response scale (either starting with the positive or the negative response category on the left side of the scale) provides information for respondents influencing their response behavior. If left means "top", then starting a response scale requesting approval of a behavior with, for example, "completely approve" or with "completely disapprove" might result in different subjective hypotheses that respondents form about the "most frequent" or "most accepted" behavior. This would imply that direction of reading is important in reacting to response scales.
If this principle applies in general, then it should be observable in response behavior of respondents in different contexts: first in a context, where reading direction - and accordingly response scales - run from left to right, and second, in contexts where reading direction runs from right to left. In the latter context, the right side of the scale would carry the meaning of "top" and response behavior can be expected to vary accordingly comparable to response behavior on scales running from left to right.
This paper tests the hypothesis that first, scales beginning with the positive pole bring about different responses than scales starting with the negative pole. Second, reading direction is important under the condition of different reading direction.
A panel design was combined with split-ballot forms of the questionnaire. Therefore it is possible to test the postulated effects of response scale direction on response behavior by comparing responses across split groups within each sample as well as across samples within each split group. Split groups of respondents did only participate once, panel members participated either twice or three times. Results of this step reveal if there is an effect of scale direction and if this depends on reading direction.