"Spatial perception, requires subjects to locate the horizontal or the vertical in a stationary display, while ignoring distracting information. Examples are the Rod and Frame Test, which requires subjects to position a rod within a tilted frame so that it is either vertical or horizontal, and the Piaget Water-Level Test, which requires subjects to draw in the water level of a tilted glass that is half filled with water. An example of the Water-Level Test is shown in Figure 3.12.
One test that is sensitive to sex differences is the "Water-Level Test" originally devised by Piaget and Inhelder (1956). In one version of this test, the subject is shown a bottle partially filled with water and is told to notice the way the water fills the bottle. The subject is then asked to predict where the water will be when the bottle is tipped. Piaget and Inhelder believed that the relevant knowledge about the horizontal would be attained at an average age of 10 years. The Water-Level Test as originally conceptualized by the developmental psychologist Piaget was never intended to test anything about water per se. It was meant to be a task of spatial concepts-in this case the ability to use a Cartesian coordinate system to represent space-but increasingly contemporary researchers discuss it as a test about the fact that the surface of water remains horizontal despite the tilt of its container, thus the meaning of this test has drifted since it was originally devised by Piaget. It seems that girls demonstrate this principle at a later age than boys. In fact, it has been estimated that 40% of college women don't know the principle that the water level remains horizontal. This is a surprising result that has been replicated many times (Wittig & Allen, 1984). Robert and Chaperon (1989), for example, reported that 32% of college women and 15% of college men failed the
"Figure A shows a bottle with some water in it.
In Figure B the bottle has been tilted.
Draw a line to show how the water line would look.
Figure 3.12 An example of the Water-Level Test. The task is to draw in the
top of the water level in Bottle B, assuming that it is Bottle A
tilted on its side."
Water-Level Test. Sex differences in the Water-Level Test have been confirmed internationally, for example, with a sample from Bombay, India (De Lisi, Parameswaran, & McGillicuddy-De Lisi, 1989). Vasta and Liben (1996) reported effects sizes that range between d= 0.44 and d = 0.66. It is difficult to understand why this should be such a formidable task for college women.
Results from the Water-Level Test are strange. Why should women (in many samples college women were used as participants) perform less well on a test of whether water remains horizontal in a tilted glass? As discussed in Chapter 6, at least part of the sex differences we find with spatial tasks can be attributed to differential learning experiences, with boys typically engaging in more spatial activities. Sex differences in the Water-Level Test are not amenable to this sort of explanation as no one believes that boys have more experiences than girls with glasses of water. In one study, Hecht and Proffitt (1995) hypothesized that experience with liquid surfaces would be associated with poorer performance on the Water-Level Test because people who work frequently with liquids in containers may have adopted a perspective that was relative to the tilt of the container-in other words they paid attention to the orientation of the container and not the level of the water surface. In a test of the hypothesis that more experience would lead to poorer performance on the Water-Level Test, Vasta, Rosenberg, Knott, and Gaze (1997) found the reverse results: subjects with more experience performed better than those with less experience. Thus, the poorer performance of females on this test remains unexplained.
Kalichman (1989) investigated the possibility that the results reflect some idiosyncrasy of the test, rather than sex differences in either the knowledge that water remains horizontal or the ability to draw an approximately horizontal line. Kalichman devised a more "ecologically valid" (i.e., more like the real world) test in which the tilted glass was held in a human hand. An example of his stimuli is shown in Figure 3.13.
Kalichman found that significantly fewer college women than college men draw an approximately horizontal line to indicate the water level in both the standard test format and in the human context format. He concluded that "sex differences on the water-level task remain robust regardless of task context" (p. 138). The Water-Level Test was used in a study of 1,704 participants ranging in age from 4 to 95 years (Tran & Formann, 2008). They used depictions of eight round bottles tilted at nine different degrees. The authors found that performance was best for adults between the ages of 16 and 60, with considerably lower performance at younger and old ages. They also reported a sex difference, which they claim was significant from adulthood into old age, but it was significant at young ages only for certain degrees of tilt. But a careful look at their data suggests that there was a floor effect at the younger ages, meaning that both girls and boys were performing so poorly at these ages that it would not be possible to tell if there was a sex difference. (See Chapter 2 for a discussion of floor and ceiling effects and how they can mislead researchers into
"Figure 3.13 The Water-Level Test embedded in an ecologically valid (i.e.,
real-world) context. The glass is half-filled with water. Draw a
horizontal line across the glass to indicate the top of the water
line. Reprinted from Kalichman (1989) with permission from the
author and Taylor & Francis."
concluding that there are no differences.) Many psychologists have studied sex differences on the Water-Level Test, perhaps because it is surprising. Correct performance on this task requires that participants understand that the surface level of water remains horizontal regardless of the tilt of the glass. As Vasta and Liben (1996) concluded in their review of this task, the puzzle is far from solved."
--- Sex Differences in Cognitive Abilities: 4th Edition / Diane F. Halpern
Damn neurosexism and patriarchy!