G. Haubensak, HOW DOES THE FREQUENCY EFFECT IN CATEGORY RATINGS ARISE - A CONTRIBUTION TO RATING THEORY, Zeitschrift fur experimentelle und angewandte Psychologie, 41(3), 1994, pp. 378-397
The same stimuli receive higher ratings when the distribution in which
they paper is positively skewed than when it is negatively skewed (fr
equency effect). The effect depends on the number of stimuli (stimulus
effect) and the number of categories (category effect). The effect ma
y be caused by one of the following tendencies: (a) the tendency to as
sign the same number of stimuli to each category (frequency model; Par
ducci, 1965; Parducci & Perrett, 1972); (b) the tendency to exaggerate
the difference between the preceding stimulus and the present one (co
ntrast model; Haubensak, 1984); and (c) the tendency to assign the sam
e categories to the same stimuli (consistency model; Haubensak 1989, 1
992a, 1992b, 1993). The three models are briefly described and evaluat
ed in the light of old and new data. The consistency model seems super
ior to its competitors in several respects.