Search results

Specific numeric values presented in a rating scale may change the meaning of the scale's verbal endpoints, & yield different results. Per 299 telephone interviews & 138 mail surveys, adults in Mannheim & Heidelberg, Germany rated 6 politicians on 2 different scales -- a "-5 to +5" & a "0-10" -- with the same verbal qualifiers of endpoints. The statistical differences in results replicated the findings of Norbert Schwarz et al (see SA 40:4/92Y9430) in that combining a verbal label with negative numeric values yielded a more negative interpretation of the verbal scale anchor, thus, a more positive movement on the scale. Results did not differ by interview method (phone vs mail). 4 References. M. Pflum

Previous research on the use of "no opinion"-filters has indicated that respondents (Rs) are least likely to offer a substantive response the more strongly the filter question is worded. Here, results of 3 experiments conducted in Urbana, Ill, & Mannheim, Federal Republic of Germany (total N = 424 college students & 336 other randomly selected adults) demonstrate that filter questions influence Rs' perception of their task: the more strongly the filter question is worded, the more Rs assume that they will have to answer difficult questions, & that they may not have the required knowledge to do so; accordingly, they are discouraged from offering global opinions. In line with this assumption, all Rs who reported not having an opinion in response to a filter question subsequently provided substantive responses on a global opinion question -- presumably because the global question asked was less demanding than expected on the basis of the filter. Analyses of these substantive responses indicate that Rs who initially reported not having an opinion differed from Rs who reported having one. Methodological implications for the use of filter questions & for research on the nature of "floating" are discussed. 3 Tables, 14 References. Adapted from the source document.

Previous survey research has demonstrated that Rs are more likely to endorse the idea that something should "not be allowed" (or "not be forbidden") than to endorse the idea that it should be "forbidden" (or "allowed"), even though these expressions seem logically equivalent. The hypothesis is advanced that this asymmetry is due to the response behavior of indifferent Rs who neither endorse that something should be forbidden nor that it should be allowed, resulting in higher endorsements of the negative form of both question wordings. A mail questionnaire administered to a random sample of 720 adults in an industrial city of the Federal Republic of Germany (87% response rate) produced data consistent with this explanation. The cognitive mechanisms underlying the response behavior of indifferents is discussed, & evidence for cross-cultural stability of the forbid-allow effect presented. 1 Table, 1 Appendix, 7 References. Modified HA

An assessment of the effects of the range of response categories provided in a closed-answer format on behavioral reports. Two sampling groups (total N = 211 West German adults) were surveyed during Mar & Apr 1983 using separate range scales to determine the amount of daily TV watched. The first group (N = 132 Rs), which had a choice of answers ranging from .5 to 2.5+ hours, reported less use of TV than the second sample (N = 79 Rs), which had a range of 2.5 to 4.5+ hours. The data on TV's importance also reflected the questions' construction, with group one indicating less satisfaction with TV but considering it more important than group two. The results demonstrate that the response scale provides Rs with a frame of reference for estimating their own behavior, & is not simply a measurement device. It is suggested that researchers consider the use of open-answer formats in obtaining data on behavioral frequencies. 2 Tables, 8 References. Modified HA