Determination of Consumer Color and Discoloration Thresholds for Purchase of Representative Retail Ground Beef

Authors:
Open Access
Publisher website
Google Scholar
Abstract
The objective of this study was to identify the threshold for color and discoloration for consumers to purchase ground beef and to determine the best objective measurement to predict consumer purchase intent. This study was designed in 2 phases, with Phase 1 requiring consumers to evaluate ground beef samples of multiple days of display simultaneously, and Phase 2 having consumers evaluate samples of only a single day of display. Ground beef packages (80% lean) were evaluated for overall appearance liking and purchase intent (yes/no) by consumers (n=216 and 318). Additionally, packages were evaluated for L*, a*, b*, calculated percentage of metmyoglobin, oxymyoglobin, chroma, hue angle, and trained sensory panel redness and discoloration scores. Models showed that each of the objective measures evaluated were predictors (P<0.05) of consumer purchasing intent. All logistic regression equations (P<0.01) had high R2 values of 0.48 to 0.86 (Phase 1) and 0.26 to 0.65 (Phase 2) and correctly classified 78.1% to 90.1% (Phase 1) and 70.5% to 84.0% (Phase 2) of samples as would/would not purchase. Linear regression equations predicting consumer overall appearance ratings with objective measures also resulted in significant (P<0.01) models, with R2 values of 0.57 to 0.93 and 0.35 to 0.54. The a* values of 21.6, 24.6, 28.3, and 30.5 (Phase 1) and 20.7, 26.2, 31.7, and 35.4 (Phase 2) correspond with consumers being 50%, 75%, 90%, and 95% likely to purchase the product at full price. However, if the product was discounted, the a* values were reduced to 17.9, 21.4, 25.0, and 27.4 (Phase 1) and 17.7, 22.7, 27.7, and 31.1 (Phase 2). The models generated from this study provide the ability to predict consumer willingness to purchase ground beef and provide ground beef processors an indication of potential consumer purchasing behaviors based upon objective values that are easy to measure.

This publication has 40 references indexed in Scilit: