[Experiences during training of a group for tasting bovine meat at an industrial refrigerated abattoir in Venezuela].

During three months, a group of people (meat consumers and active personnel) of a Venezuelan packing house were asked to assign discretionary scores to a sample of 103 beef longissimus steaks in order to familiarize them with the score sheet and to their future job as panelists. Scores assigned to 37 steaks by seven candidates that assisted regularly to these preliminary sessions (out of 12 attendants), were subjected to Analysis of Variance (ANOVA) to determine variability among judges before training. The whole group (n = 12) potential candidates from both sexes, and ranging between 25 and 45 years, were interviewed and pre selected for formal training. Triangle tests were used as selection tool and allowed for dismissing two of them. Training reinforced acquaintance with test procedures and improved natural abilities, sensitivity and memory for test attributes by using an 8-point descriptive scale to describe samples differing in fiber tenderness, amount of connective tissue, overall tenderness, juiciness, and flavor intensity. After 2 mo. of training (5 h weekly) the first panel performance evaluation (PPE) was conducted over 4 days with 3 sessions per day and 3 samples per session with 9 different samples. Data for each candidate were subjected to one-way ANOVA and F-ratio served as criteria to rank candidates' ability for describing each attribute. Panelist 7 and 8 rated last and were excused to conform an 8-member panel. The panel evaluated 300 samples during a 6-mo period and a second PPE was conducted. The second PPE revealed changes in relative ranking positions. Data from the first 40 (out of 300) samples were subjected to ANOVA for testing score (panelist) variability after training. A significant effect of panelist was detected before (P < .05) and after (P < .001) training, but a much greater variability became apparent in untrained candidates. Mean separation for trained panelist only differed in decimal units within the same points or adjacent points of the scale. Results validated methodologies used for panel training and the panel capability for sensorial evaluation of non-processed beef.