Dot displacements can be deceptive: a reply to Hartley.

Abstract

Hartley (1979) has replied recently to my comments (Wenderoth, 1978) on an experiment and an analysis of his (Hartley, 1978). One of the points I had made was that when Hartley had observers place dots on blank sheets of paper to match the apparent positions of. dots within outline frames, these judgments mirrored the appearances of the surrounded dots. That is, if dot appeared too far to the left within frame, the subject would indicate this by placing dot too far to the left on the blank sheet. In my experiments, the method of measurement was different: subjects had to set surrounded rod to apparent vertical. It needed to be pointed out that in the latter case, rod settings compensated for the appearance of truly vertical rod, rather than representing it. Hartley had stated (see Wenderoth, 1978) that the two experiments gave opposite results; I wished merely to note that the results were logically equivalent, rather than opposite, as he wrongly believed. In his recent reply, Hartley (1979) suggests that when dot within, say, square frame is closer to the center than to the sides of the square, the major (Wenderoth, 1978)predicts that the dot will be repelled from the nearer virtual axis of symmetry rather than the real line (side) and so will appear to be further from the square's center than it really is. He claims that the major axes account can predict this result, because now he accepts that appearances and compensatory settings are directionally opposite. But, he notes, directionally similar errors also occurred in his experiment when the enclosed dot was nearer the frame's side than its center. This, he asserts, is contrary to the major axes prediction. It needs to be reemphasized (see Wenderoth, 1978) that the major axes hypothesis refers to the orientation domain. That is, the hypothesis asserts that nonvertical frame will induce an apparent tilt in vertical rod and that this induced tilt will be in the direction away from whichever frame axis of symmetry is nearest in orientation, not position, to the rod. The hypothesis has never referred, or been extended, to the position domain, which is why I concluded earlier (Wenderoth, 1978, p. 574) that a test