Partial least squares (PLS) regression for the analysis of instrument measurements and sensory meat quality data

The partial least squares (PLS) regression technique was used to examine meat quality data derived from instruments (including Warner-Bratzler shear force and Instron Compression) and sensory panels. The data related to beef longissimus dorsi muscles collected during trials to study the effect of hot boning on meat quality. The univariate analysis of tenderness showed that over 60% of the variation in sensory tenderness, and almost 60% of the variation in sensory acceptability, could be explained from instrument variables and a consideration of boning and ageing time. Graphical displays from the analysis indicated that hot boning (either at 1 or 4 h) had little effect on meat quality. Graphical displays demonstrated a possible important effect of vacuum-pack ageing on acceptability. For this data set, it appears that samples of approximately equivalent tenderness differ in acceptability, depending on whether the samples have been aged for 1 or 4 weeks. This finding may have practical importance in attempts to predict eating quality (acceptability) from instrument measurements. Separate equations are necessary for products aged for different periods. (C) 1999 Elsevier Science Ltd. All rights reserved.