Demographic and design effects on beef sensory scores given by Korean and Australian consumers

摘要

Data from 648 beef samples, which had been sensory tested by 720 Korean and 540 Australian consumers were used to quantify design and demographic effects on beef sensory scores. The samples were from 36 carcasses, where sides had been either hung by the Achilles tendon or hip suspended. At boning, samples from three muscles (M. triceps brachii, M. longissimus dorsi and M. semimembranosus) were prepared and cooked by either grill (25-mm-thick steaks) or Korean barbeque (BBQ, 4-mm-thick samples) methods. A Latin square design was used to allocate samples to different presentation orders to be tasted in association with different samples. For both cooking techniques each consumer tested a starter sample followed by six experimental samples, with each sample being tasted by 10 different consumers.nnDesign (taste panel, session, order, carry-over, sample and consumer) and demographic (age class, gender, occupation, frequency of eating meat, number of adults and children living in the house, their appreciation of meat and degree of doneness and income) effects were examined separately for tenderness, juiciness, like flavour, overall liking and a composite palatability score, within the four consumer group/cooking method subclasses. For grill samples, order of presentation was significant for most sensory variables. For BBQ samples, order of presentation failed to achieve significance for Australian consumers, but was significant (P < 0.05) for Korean consumers. Carry-over effects tended to be more important for juiciness and like flavour scores than other sensory scores. Demographic effects were generally not significant (P > 0.05) for all consumer group/cooking methods. Correlations between raw scores and those adjusted for design and demographic effects ranged from 0.93 to 0.99, indicating that if the design was balanced, or nearly balanced for design effects, then further adjustment of sensory scores was not necessary. Clipping 40% of outlying consumer scores reduced the variance of the sample mean by ~30%.