A masking approach to the problem of form perception

The effects of deletion of certain components of a dotted triangle are studied, using a masking technique. Masking interference is achieved by varying the density of randomly placed dots within which stimulus triangles must be detected. The results of the study indicate that deletion of the dots composing the corners of the triangles decreases performance no more than deletion of an equivalent number of randomly chosen dots. However, the detectability of the line of dots composing the sides is functionally related to the recognition of the orientation of the triangle. The results suggest that the relationship of the parts of such a stimulus figure may be more important than the specific parts themselves.