Response scales as frames of reference: The impact of frequency range on diagnostic judgements
✍ Scribed by Norbert Schwarz; Herbert Bless; Gerd Bohner; Uwe Harlacher; Margit Kellenbenz
- Publisher
- John Wiley and Sons
- Year
- 1991
- Tongue
- English
- Weight
- 833 KB
- Volume
- 5
- Category
- Article
- ISSN
- 0888-4080
No coin nor oath required. For personal study only.
✦ Synopsis
In social and psychological research, respondents are often asked to report the frequency of a behaviour by checking the appropriate alternative from a list of response categories provided to them. Previous research indicated that respondents extract comparison information from the range of the response alternatives, assuming that the average respondent is represented by values in the middle range of the scale, and that the extremes of the scale represent the extremes of the distribution. Extending this line of research, the present studies demonstrate that the users of a respondent's report are also likely to use the range of the response alternatives as a frame of reference in evaluating the implications of the report. Specifically, subjects are found to draw different conclusions about the respondent's personality (Experiment l), or the severity of his or her medical condition (Experiment 21, from the same absolute frequency report, depending upon the range of the response scale on which the frequency was checked. Moreover, experienced medical doctors were as likely to be influenced by scale range as first-year medical students, suggesting that the phenomenon is of considerable applied importance. Implications for the use of response alternatives in psychological research and diagnostic judgement are discussed.
In psychological testing, as well as in laboratory experiments and survey research, respondents are often asked to report the frequency with which they engage in a certain beaviour or make a certain experience. To obtain the desired behavioural information, respondents are typically asked to check the appropriate alternative from a set of response categories provided to them. The selected alternative is assumed to inform the researcher about the respondent's behaviour. It is frequently overlooked, however, that a given set of response alternatives may be far more than a simple 'measurement device'. Rather, it may also constitute a source of information for the respondent, because respondents assume that the range of the response alternatives reflects the researcher's knowledge of, or expectations about, the distribution of the behaviour in the 'real world'. Specifically, they assume that the average behaviour is represented by response alternatives in the middle range of the scale and that the extremes of the scale reflect the extremes of the distribution (see Schwarz, 1988, in press for reviews).
Accordingly, respondents were found to extract comparison information from the range of the response alternatives provided to them (Schwarz, Hippler,