Sensory descriptive analysis and correspondence analysis aids in the choosing of apple genotypes for processed products

A trained sensory panel developed a descriptive vocabulary and procedures t o evaluate the fruit of 30 apple genotypes processed as apple pies. The sen sory attributes included seven terms to quantify color and appearance, seve n for flavor, and eight for texture. A generalized lattice design was used to select subsets of the genotypes for evaluation in the panel sessions. Ge notype means were estimated for each descriptive term using Residual Maximu m Likelihood (REML), from which sensory profiles were generated. Correspond ence analysis was used to define distinct components of the sensory profile s, and then to select genotypes that were similar in sensory properties to the standard industry genotype for pie processing, Northern Spy. Correspond ence analysis portrayed the principal associations among the 22 sensory ter ms and 30 apple genotypes in four dimensions. The combined sensory and stat istical techniques revealed that one-fifth of the apple selections have pot ential for processed products.