Rating Visual Velocities: Presentation Order versus Frequency of Occurrence

Abstract

Given equal frequencies of occurrence of stimuli, identical velocities receive higher or lower ratings depending on whether slow or fast velocities happen to occur on the early trials of a run (Ehrenstein and Sokolov, 1996 Perception25 Supplement, 34 – 35). When stimuli vary in their frequency of occurrence, this presentation-order effect might enhance or substitute the well-known frequency effect (Parducci's range - frequency theory) on category ratings. The aim of the present study was to contrast these two effects for ratings of visual velocities. Two groups of observers had to rate five stimulus velocities (3, 4.5, 6, 7.5, and 9 deg s−1) of a single luminous dot using three categories (slow, medium, fast). Respective frequencies of occurrence for these velocities were either 4 - 4 - 8 - 14 - 20 (group 1, negative skewing for frequency) or 20 - 14 - 8 - 4 - 4 (group 2, positive skewing). The quasi-random presentation order of each set, on the contrary, corresponded to either positively skewed (group 1) or negatively skewed (group 2) frequency distributions. No significant differences in the ratings of two groups (other than a slight preference for the presentation-order effect beyond 6 deg s−1) were found, as if the presentation order and frequency effects had cancelled each other. This suggests that the presentation order of velocities is at least as powerful as their frequency of occurrence. Lack of agreement between our results and predictions of computer-simulated Haubensak's consistency model as well as of Parducci's range - frequency theory requires a model which accounts for an interaction of the two effects.