The left-right self-placement is one of the most frequently used measures for ideological self-identification in empirical political science research. However, the respondents’ understanding of “left” and “right” is only rarely tested.
Recent research with the German General Social Survey (ALLBUS) 2008 data has shown that a considerable amount of respondents who place themselves on the left-right scale do not answer if asked afterwards about the meaning of “left” and “right”. While this might question the validity of the left-right scale, there might be several reasons for non-response on the open-ended question. A reason might be that respondents are not offered an explicit “can’t choose” category in the left-right scale which then results in “forced” placement on the left-right scale. Methodological research has demonstrated that respondents without opinion but forced to answer mostly choose the middle scale values.
In a web survey based on an sample from a German online access panel we tested to what extent an explicit “can’t choose” category as well as different scale formats (10-point vs. 11-point scale) of the left-right self-placement affect the results with regard to the left-right scale itself and also with regard answering the open-ended questions.