The Training and Selection of a Panel for the Sensory Evaluation of Beef

The training and selection of persons for a panel for sensory evaluation of beef proceeded in three phases - - preliminary screening by triangle tests, training sessions with hamburger patties with known differences in composition, and training with 6-7-8th rib roasts varying in USDA grades. Various statistical criteria were used to rank panelists' performance on scoring flavor, juiciness, and tenderness of hamburger patties and roasts. These criteria were the percentage of correct responses on the triangle tests; R² values and F ratios for flavor, tenderness, and juiciness scores and correlation coefficients for juiciness scores versus percent fat in cooked hamburger patties; and correlation coefficients for juiciness scores versus numerical USDA grades and for tenderness scores versus Warner-Bratzler shear values from the roasts. Panelists were ranked according to each of the above criteria and mean ranks with standard deviations and rank correlations were computed.

Results indicated that the triangle tests were not a good screening device for prospective panelists. Judges varied in their ability to maintain the same rank order in scoring different attributes of one product or similar attributes in two beef products. New judges for the panel on sensory evaluation of beef were selected on the basis of mean ranks.