CONTEXT EFFECTS AND THEIR INTERACTION WITH DEVELOPMENT - AREA JUDGMENTS

The perceptual constancy view posits that children perceive rectangular areas dependent on shape, whereas adults do not. Following Brunswik (1934a), we call this dependency the "limited perceptual constancy of area under variation of shape." A series of four experiments (with 150 children and 80 adults) using ordinal paired comparisons rejected the general equation ai = (Ai, Si) as a model of limited area constancy (ai is the perceived area, Ai and Si the physical area and shape of rectangle i). Children judged area dependent on shape, but dependent on the covariation Cset of area and shape in the experimental set of rectangles, rather than on the particular shape of the rectangle judged. A quantitative version of this covariation mechanism that seems to generate limited area constancy is proposed, and its relation to Parducci's range-frequency theory is discussed. Not one child's judgments could be predicted from a height + width rule (assuming identical linear psychophysical functions). We demonstrate and correct serious errors in Anderson and Cuneo's (1978a, 1978b) claims for a height + width rule in children. Our results support a three-step developmental model for paired comparisons of area: (1) centering on one side of the rectangle, (2) limited area constancy, that is, ai = f(Ai, Cset), and (3) area constancy. © 1990.