Illusory Tilt and Euclidean Schemes as Factors in Performance on the Water-Level Task

Predictions made by models of water-level acquisition were tested to understand better the sensory factors, cognitive factors, or both, that differentiate people who fail the test (low scorers) from people who pass (high scorers). Experiments 1 and 2 showed that in a horizontal-edge detection task, low scorers were less likely than high scorers to "see" liquid edges in tilted containers as horizontal. Experiment 3 showed that water-level type displays belong to a class of tilt illusions that arise from early visual processes for both low and high scorers, but only high scorers spontaneously use cognitive schemes to minimize illusory tilt. In Experiment 4, high scorers, but not low scorers, overrode the orientation illusion in a production task. The findings suggest bottom-up processes cause embedded lines to be misperceived for both groups, but high scorers are more likely to use cognitive schemes spontaneously to overcome the orientation illusion.