Comparative judgment: tests of two theories using the Baldwin figure.

Two theories of comparative judgment were compared across four experiments on their ability to explain the Baldwin figure, a focal line whose apparent length is affected by square boxes at or near its endpoints. Left-box size, right-box size, line length, line-box distance, and other variables were varied in factorial designs to allow application of functional measurement methodology. The model from adaptation level theory did poorly in several respects. In particular, it had trouble with the pervasive lack of contrast. Further, it could not account for the fact that a box added contralaterally increased the illusion, whereas the same box added ipsilaterally decreased the illusion. The model from information integration theory did substantially better, though it too had trouble with some of the results. An alternative interpretation was suggested in which the Baldwin figure is viewed as a positive context or assimilation effect. This positive context formulation may generalize to other illusions, such as those produced by the Müller-Lyer and Ponzo figures.