Judgments of contour strength or saliency for twenty-four illusory-contour configurations were subjected to a confirmatory factor analysis. A four-factor model that posited the involvement of simultaneous contrast, linear effects (assimilation and dissimilation), depth/completion cues, and feature analyzers accounted for a substantial proportion of the variance in judgments of illusory-contour strength. The hierarchical addition of a fifth factor, diffuse illusory contours, significantly improved the overall fit of the model, but added little to the proportion of explained variance. The taxonomic approach adopted provides support for a multiprocess model of illusory-contour perception.
Independence of the completion effect from the noncompletion effect in illusory contour perception
