Taste thresholds for sodium chloride in longitudinal experiments.

Abstract

A rapid ascending method of limits without standards was reliable for determining taste thresholds for NaCl in longitudinal experiments. Three experienced Ss made two judgments—tentative-uncertain and certain (absolute threshold) in each of five series of taste judgments daily for 120 consecutive days. To provide a means of studying functional relationships between taste and other variables, the total variance in all the judgments was partialled into several orthogonal components by means of linear components analysis. Factor scores computed from these components were correlated with a series of values derived directly from the judgments. The first component for each S consolidated the variance that arose from the average day-to-day differences in judgments, and the others isolated the variance that was due to within-day linear trends. The relation of the unrotated and rotated components was shown, and the reasons for our preference for the former were explained. The factor scores from these several components of variance may be viewed as new variables that measure these structural pools of variance and, as such, they may be correlated with any other pertinent variables such as constituents of saliva, motivation, environmental temperature, and the like. Furthermore, the estimation of the “subliminal” threshold not only was reliable, but it expands the possibility for clarification of the nature of taste percepts.