Metacontrast: Internal contours and different dependent variables

Abstract

Extent of metacontrast was assessed using square-wave black and white gratings as targets. The number of internal contours in the target and delay of the mask were systematically manipulated. Three experiments each used a different dependent variable; Experiment 1 used accuracy, Experiment 2 used duration threshold, and Experiment 3 used category estimations. Results showed, (1) targets with many internal contours are difficult to see, (2) there were no backward enhancement effects, (3) targets with few contours and targets with many contours, when detected, are relatively easy to mask, (4) targets with intermediate numbers of contours (2.54 to 8 cycles/deg) are relatively easy to see and produce monotonic metacontrast functions, but, (5) category estimation measures showed some evidence of U-shaped metacontrast functions. The results are discussed in terms of task differences.