Objective Evaluation of Canned Tuna Sensory Quality

Abstract

SUMMARY— Commercial and experimental packs of canned tuna were prepared from the 1965, 1966 and 1967 packs. The 1965 samples were used to train an expert panel which identified 8 distinct parameters of quality. Of the 64 objective tests or variations of test procedures, 13 were selected for further study. 1966 analyses revealed that color, flakiness, firmness and fiber could be predicted adequately (R = 0.9 +). The 4 other parameters of scorch, juiciness, odor and taste were also highly significantly correlated, but were predictable less completely (R = 0.8 – 0.9) by 1 or a combination of objective tests. 1967 samples were submitted to a consumer preference panel in addition to measurements by expert panel and objective tests. These results indicated that consumer preference is dependent approximately equally (40%) on appearance and flavor and to a lesser extent (20%) on textural characteristics. Consumer preference could be estimated objectively by determinations of Hunter L and b values, a shear test, percent fiber and PH.

The role of the human senses in food acceptance

“…One common approach is to conduct a multiple regression of individual sensory attribute ratings against overall liking (e.g. Rasekh et al, 1970;Powers et al, 1977;Azanza et at., 1994). If the attribute ratings have been normalized, the obtained coefficients of the equation will index the degree of importance of the attribute to overall liking.…”

Muscle biochemical constituents of cultured amberjack Seriola dumerili and their influence on raw meat texture

“…The most important quality attribute of animal and fish products-from a consumer's point of view-are sensory attributes, such as colour, texture, odor, and flavor [1]. Eating quality is also an important determinant of the overall impression of a food product [2], and the eating quality of a fish is determined by its composition. Kestin et al [3] showed that the muscle lipid level significantly affected the eating quality of rainbow trout.…”

Nonparametric Statistics: Advanced Computational Methods

“…The interval [a n , b n ] is an approximate 1 -2a confidence interval for s. Standard theorems show that the confidence level of this confidence interval converges to 1 -2a as K, n / N. Rasekh et al (1970) wish to evaluate the relative importance of factors contributing to tuna quality, in particular the importance of the Hunter L value factor. This H n is the empirical distribution function of the bi-variate sample.…”