Quantification of Sensory and Food Quality: The R‐Index Analysis

Abstract

The accurate quantification of sensory difference/similarity between foods, as well as consumer acceptance/preference and concepts, is greatly needed to optimize and maintain food quality. The R-Index is one class of measures of the degree of difference/similarity, and was originally developed for sensory difference tests for food quality control, product development, and so on. The index is based on signal detection theory and is free of the response bias that can invalidate difference testing protocols, including categorization and same-different and A-Not A tests. It is also a nonparametric analysis, making no assumptions about sensory distributions, and is simple to compute and understand. The R-Index is also flexible in its application. Methods based on R-Index analysis have been used as detection and sensory difference tests, as simple alternatives to hedonic scaling, and for the measurement of consumer concepts. This review indicates the various computational strategies for the R-Index and its practical applications to consumer and sensory measurements in food science.

Effects of vitamin E and fish oil inclusion in broiler diets on meat fatty acid composition and on the flavour of a composite sample of breast meat

“…Sensory analytical data were collected using the Compusense  data capture system (Compusense Inc., Guelph, Canada). The R-index of each test sample relative to the control was estimated 10 and compared with the critical R-Index at P < 0.05 12 to determine if the samples were significantly different from the reference (0LC100E sample) at P < 0.05.…”

“…10 This is an estimate of the probability that the breast meat produced from birds fed fish oil may be identified as being different from the reference (0LC100E) meat. The R-index ranges in value from 50% (where a judge cannot discriminate between the sample and reference meat, and so has to guess whether they are different, and therefore has a 50% chance of www.soci.org C Rymer, DI Givens being correct) and 100% (where a judge can discriminate perfectly between the two meats).…”

Influence of the fiber from agro-industrial co-products as functional food ingredient on the acceptance, neophobia and sensory characteristics of cooked sausages

“…In meat products, R-index values greater than 50% indicate a higher hedonic rating (Pasin et al 1989). Added pineapple fiber had no effect on sausage firmness (R-index = 50%), whereas the higher R-index value close to 100% (90.29%) indicated that samples formulated with cactus pear fiber were clearly preferred by judges (Lee and Van Hout 2009). For taste, R-index values were slightly higher for pineapple fiber samples than for cactus pear fiber (83.75 and 71.24%, respectively), with a slight preference for sausages containing pineapple fiber.…”

Equivalence tests – A review

“…Bi (2005) specified the equivalence test problem and discussed several approaches. Applications of equivalence tests were proposed for different situations like analysis of the R-index (Bi, 2006;Lee & van Hout, 2009), d 0 as determined from forced-choice methods (Bi, 2011a), in the framework of Bayesian models for difference and preference tests including the evaluation of equivalence hypotheses Bi (2011b), and for unreplicated paired comparisons (Bi, 2007a). For replicated paired comparisons, Bi (2007b) proposes an approach based on the Beta-binomial model, using the overdispersion parameter as a nuisance parameter only without defining explicit margins for it and therefore unlike the approaches discussed in the previous section.…”