An Adjustment Error in Optimization of Taste Intensity

In a currently popular method of adjustment for assessing optimal preferred levels of tastants in foods, a robust perceptual error was observed. When subjects concentrated a dilute beverage to optimal taste, the final (adjusted) concentration of the added tastant was reliably lower than when diluting a concentrated beverage. This discrepancy between ascending and descending runs was observed for both adjusted sucrose in a fruit beverage and for adjusted NaCl in tomato juice. The premature shift in perception from non-optimal to optimal judgment resembles the “error of anticipation” in the classical method of limits. The effect could not be attributed to sensory adaptation or perceptual insensitivity, and it persisted even when subjects were told to be sure to reach the same level in both directions of adjustment. The magnitude of the error was diminished by interweaving of ascending and descending trials, which altered the immediate stimulus context, and by enhancing the motivational level of subjects (through reward) to try to reach the same concentration in both ascending and descending adjustments.