Subjects were presented five concentrations of sucrose and water in a randomized order and recorded responses on either the “sweet” target or “water” target in the response field of the touch-sensitive display. All “correct” responses (touching the “sweet” target on trials of any concentration of sucrose or the “water” target on trials of water) were reinforced, and “incorrect” responses were penalized (as described in Materials and Methods). The value for discriminability, d’, which takes response bias into account (see Data Analyses), was calculated from the cumulative results of 27 tests (13 subjects tested twice; one subject tested once).