To our knowledge, there has been only one study that tested children’s imagery abilities using the
water tilting task (Black & Schwartz, 1996). In this experiment, the diameter of the glasses and the level
of the imaginary water inside the glasses were varied on two levels each in a complete factorial
design. Unfortunately, the results of this study were described only briefly in a short proceedings paper.
They showed that most 3- to 12-year-olds correctly tilted the thin glasses farther than the wide
ones, with the exception of 5-year-olds, when tilting glasses with low water levels. The water levels
were discriminated only by older children (at least 8 years of age). The 3- and 4-year-olds correctly
discriminated the diameters but not the water levels. This result is rather counterintuitive in that
the water level would seem to be the ‘‘easier” factor for which we have more accurate descriptive
knowledge. The authors did not discuss this issue. Based on the results that at least older children
(starting at around 8 years of age) discriminated both relevant factors even though they did not show
any descriptive knowledge about the effects of diameter on the tilting angle, the authors concluded
that the children solved the task by using dynamic imagery.