web analytics

Aspects of perception of tenderness of beef

Back to News & Events
May 10, 2011

This study evaluated temporal differences amongst panelists in perception of juiciness and tenderness of meat. Use of Principal Component Analysis (PCA) to produce curves based on PC scores over time provided more information about the samples and perception variability than simple averaging.


The TI method, by having panelists continuously monitor their perceived sensations, offers a unique advantage over conventional methods of texture measurement by showing the temporal aspects of texture perception. The information obtained is expressed as curves representing intensity over time, facilitating inter-sample comparisons. However, the TI method, which records data at frequent intervals, is even more susceptible to individual influences than traditional single point scaling methods. Ignoring differences between individual curves is a common practice in TI analysis resulting in loss of temporal information. The purpose of this research was to evaluate temporal differences amongst trained panelists in perception of juiciness and tenderness of beef muscle. To do this the usefulness of PCA in dealing with individual variability and in separating different patterns of temporal perception within a group of trained panelists was explored.


On the basis of typical tenderness curves generated for each panelist by the Compusense software, panelists were grouped by the visual inspection into three groups. These could be likened to the panelists ‘chewing patterns. The largest group, containing 5 panelists, was strong efficient chewers masticating the meat in a short time. The second group, consisting of 3 panelists, perceived a greater force required to chew for a longer time after obtaining maximum force (Imax), but were also efficient chewers. The last group of 2 panelists was considered to be inefficient chewers as they perceived a large amount of force being required to chew the samples at Imax for considerably longer time than the other two groups. Typical juiciness curves, generated for each panelist, grouped the panelists into four groups for perception of amount of juiciness over time. Four of the 10 panelists, showed a sharp increase in juiciness with a quick decline, which was probably influenced by their chewing style. Other three panelists exhibited variations on a plateau effect for maximum juiciness perception, whereas another two panellists showed increasing perceptions of juiciness by a series of small plateau followed by a sharp decline. One panelist exhibited inconsistency in juiciness perception. Juiciness perception was influenced by the amount of breakdown caused by mastication at a given point in time. Also eight of the 10 panellists exhibited, to some extent, a ‘ski jump’ effect at the end of the juiciness evaluation. This was caused by the fact that juiciness, unlike tenderness, persisted throughout the mastication to the point of swallowing and thus terminated abruptly.

Panelist performance for tenderness and juiciness assessment was examined by PCA.
The First and Second Principal Components explained most of the variability (84%) in the original ten panelist data matrix. The plot of the loadings of the individual panelists provided good information about panel homogeneity and indicated where differences between panelists occurred. Construction of profile curves using First and Second PC scores for individual panelists, and for the panel, was useful for grouping panellists with similar chewing behaviour. Using PC scores rather than simple averaging techniques for TI data was found to be particularly useful where panelist differences occurred such as assessment of juiciness and force required to chew less tender meat samples. First PC curves were useful in examining how perceptions of one attribute could induce perceptual differences in the other. The possibility exists that PC information regarding chewing behaviour of individual panelists, combined with the information regarding the perceptual interrelationships between tenderness and juiciness, could be explored as to the effect on acceptability. Not only would this provide a better understanding of acceptability of meat but it should also make it possible to use trained panel data as an indicator of acceptability.