Comparison of Preference Mapping Methods on Commodity Foods with Challenging Groups of Low-Variance Attributes: Sliced Whole Wheat Sandwich Bread Example

Abstract

Listen

Translate article

Traditional preference mapping methods can suffer where groups of attributes with larger variances dominate the analysis, thus, detracting attention from attributes of potential importance as drivers of liking. This study compared traditional external preference mapping methods (PLS and PREFMAP) with a new method called PrefHMFA which is designed to control dominance of groups of high variance attributes. Twenty-five sliced whole wheat breads were profiled by descriptive analysis (DA) for flavor, appearance (crust and crumb groups) and texture (oral, hand) attributes. Breads were subsequently presented to adult bread consumers (n = 360). Data were also subjected to path analysis (PATH-PLS) and three different preference mapping analyses (PLS and PREFMAP and PrefHMFA). Traditional methods (PLS/PREFMAP) showed broad ideal points. PrefHMFA partial axes showed that the main hedonic dimensions were aligned with higher sensory dimensions. PrefHMFA revealed the greater importance of appearance and hand-perceived texture attributes for liking. Path-PLS confirmed the importance of sensory liking and family drivers of purchase intent for sliced sandwich bread. Practical Applications External preference mapping is used to relate analytical sensory data to consumer acceptance. Traditional techniques have been criticized since groups of attributes with large variances can dominate the traditional PCA or PLS-based methods. New preference mapping techniques of PrefMFA and PrefHMFA have been suggested to account for attribute dominance. These techniques will aid in determining drivers of liking for commodity-type foods and other product or situations where attribute dominance poses a problem with traditional methods.