Relationship between Witkin's Rod-and-Frame Task and Piaget's Water-Line Task for College Women

Abstract

The water-line concept is the understanding that a liquid surface is invariantly horizontal, regardless of how the container of the liquid may be tilted. Piaget and Inhelder ( ) report that children do not develop this concept until about the age of nine and their observations have been verified by others (1, 3, 4 ) . Thomas (4 ) and Willemsen and Reynolds ( ) have found that approximately half of college women cannot correctly represent the water line in a tilted container whereas nearly all college men can. Willemsen and Reynolds (5) have shown that the deficient performance of some women on the water-line task is due to their greater tendency, relative to men, to be distracted by the straight-line characteristics of the sides, top, and bottom of the container. This tendency to be guided by external or irrelevant features of a perceptual situation is also shown by adult women on the Rod-and-frame task ( 6 ) . This study was undertaken to estimate the correlation between the two tasks with a view to developing an understanding of sex differences on the water-line task in terms of Witkin, et al.'s concept of differentiation (6 ) . Ss were 30 undergraduate women enrolled at a small Catholic university in the West. Three undergraduate psychology majors (A.B., M.S.B., and N.G.) served as Es. Ss were assigned to them at random. Half the Ss were given the waterline task first and the Rodand-frame task second, and half had the reverse order. For the water-line task, the Thomas Mark I apparatus as described in Willemsen and Reynolds ( ) was uesd along with a cylindrical container. This apparatus permits assessment of S's ability to represent the water line on a tilted container correctly. Twelve trials were conducted on this task-one for each angle of tilt corresponding to an hour on the clock. S's representations of the water line were scored as the average absolute number of degrees from the horizontal. Twelve trials on the Rod-and-frame task were given on the standard table top model with a Min. square frame manufactured by Shaw Laboratories, Inc., Syosset (L . I . ) , New York (Model PR-20). The frame settings varied in 5 increments from 25 left of vertical to 25 right of vertical. Rod settings varied independently in 2 increments from 7 left or right to 25 left or right. S was instructed to set the rod to the vertical and her error score was taken as the average absolute number of degrees from the vertical of her settings. As a non-parametric Kruskal-Wallis analysis of variance by ranks indicated no significant differences among Es in the error scores on each of the two tasks, data for all 30 Ss were combined. The mean error on the Rod-and-frame task was 2.86, SD = .83. The mean error for the water-line task was 4.31°, SD = 6.08. A Spearman rho between errors on the two casks was .37 ( r > 0; f i = .05). This modest value suggests that lack of differentiation as assessed by the Rod-and-frame task may account for some of the variance of scores on the water-line task but that other variables contribute as well.