Detection vs. grouping thresholds for elements differing in spacing, size and luminance. An alternative approach towards the psychophysics of Gestalten

Abstract

Three experiments were performed to compare thresholds for the detection of non-uniformity in spacing, size and luminance with thresholds for grouping. In the first experiment a row of 12 black equi-spaced dots was used and the spacing after the 3rd, 6th, and 9th dot increased in random steps to determine the threshold at which the observer detected an irregularity in the size of the gaps. Thereafter, spacing in the same locations was increased further to find the threshold at which the observer perceived four groups of three dots each (triplets). In the second experiment, empty circles were used instead of dots and the diameter of the circles in the first and second triplet increased until the difference in size gave rise either to a detection or grouping response. In the third experiment, the dots in the second and fourth triplet were increased in luminance. The aim again was to compare the difference in brightness required for detection or grouping, respectively. Results demonstrate that the threshold for perceiving stimuli as irregularly spaced or dissimilar in size or brightness is much smaller than the threshold for grouping. In order to perceive stimuli as grouped, stimulus differences had to be 5.2 times (for dot spacing), 7.4 times (for size) and 6.6 times (for luminance) larger than for detection. Two control experiments demonstrated that the difference between the two kinds of thresholds persisted even when only two gaps were used instead of three and when gap position was randomized.