The objective of this study was to determine the accuracy and repeatability of untrained laboratory consumer panelists in detecting differences in beef longissimus tenderness. At 14 d postmortem, slice shear force was measured on one steak from 192 strip loins and used to select 54 strip loins and assign 18 of the strip loins to each of three tenderness classes (tender = <15 kg, intermediate = 15 to 27 kg, and tough = >27 kg). Sixty-eight untrained, laboratory consumer panelists evaluated paired steaks from each tenderness class in each of two sessions (12 total observations per panelist). Mean slice shear forces for "tender," "intermediate," and "tough" were 11.1, 21.0, and 32.2 kg, respectively. Mean tenderness ratings of the untrained laboratory consumer panel were different (P < 0.05) among tenderness classes (mean of 16 panelists = 6.2, 4.9, and 3.3 for tender, intermediate, and tough, respectively), and these differences were similar regardless of how many untrained panelists were averaged to determine the panel mean (4, 8, 12, or 16). The correlations (P < 0.01) between slice shear force and the mean untrained consumer panel tenderness rating (mean of 4, r = -0.82; mean of 8, r = -0.89; mean of 12, r = -0.91; and mean of 16, r = -0.92;) were similar. Overall repeatability of the untrained consumer panel was 0.80. Repeatability of individual untrained consumer panelists for tenderness rating was highly variable: 31% were >0.80, 36% were 0.60 to 0.79, and 33% were <0.60. Thirty-two percent of the consumers were both accurate (correlation to slice shear force = -0.75 to -1.00, P < 0.01) and repeatable (repeatability >0.75). There is wide variability in the ability of untrained laboratory consumer panelists to detect differences in beef tenderness. Nonetheless, untrained consumer panels can accurately and repeatedly detect differences in beef tenderness under controlled laboratory conditions. An untrained laboratory consumer panel may be able to provide as effective an evaluation of beef longissimus tenderness as a trained descriptive attribute panel.
