| Abstract: | Frequency of behaviour is often assessed by scales using relative frequencies such as ‘often’ or ‘rarely’, so‐called vague quantifiers. Previous research showed that respondents calibrate such scales according to subjective standards. Here, it is argued that respondents follow conversational norms and if possible try to figure out which calibration the researcher had in mind and adapt their responses accordingly. They may use the survey context to infer a relevant anchor for such vague quantifiers. In the present study, although respondents did not differ in absolute behaviour frequencies, their reports of relative frequencies were influenced by information about the target population and the topic of the survey. Apparently, respondents anchored the scale according to the estimated frequency in the target population and the frequency of other behaviours relevant to the topic of the survey. Copyright © 2002 John Wiley & Sons, Ltd. |
