The effects of verbal and numerical quantifiers on questionnaire responses

It is concluded that numerical quantifiers are probably more useful than verbal ones, and should be used if possible.

Abstract

Three experiments are described which assess the effects of responding to different sets of verbal quantifiers (e g, always, often, sometimes). In Experiment 1, participants were asked to give percentage equivalents to three sets of supposedly equivalent quantifiers. The responses differed significantly. In Experiment 2, participants completed different versions of a questionnaire in order to see if the three sets of quantifiers produced equivalent responses in a more concrete situation. The results again showed that the three sets were not equivalent. In Experiment 3, participants completed one version of the questionnaire with a set of verbal quantifiers and two versions of the questionnaire with numerical quantifiers (e g, I do this about 75% of the time). The verbal quantifiers produced virtually the same results as before: the numerical quantifiers produced different but quite consistent results. It is concluded that numerical quantifiers are probably more useful than verbal ones, and should be used if possible.