Comparison study of sensory and instrumental evaluation of bread texture using PLS regression analysis

In this study, Partial Least Squares (PLS) statistical procedure was used to analyse the results of texture measurements of gluten-free breads prepared from different proportions of pectin, guar and carageenan hydrocolloids. Correlation coefficient between instrumental firmness of bread with sensory quality attributes of resilience, porosity, chewability and staling were found to be 0.75, - 0.75, -0.40 and -0.71, respectively. Bread sensory attributes showed good correlations with quality attributes such as loaf volume and height, indicating that breads with higher loaves and heights corresponded to softer crumb and better crumb resilience. The results obtained for VIP coefficients in PLS regression for instrumental firmness, loaf volume and height variables indicated that instrumental firmness could able to better predict the individual sensory attributes of bread. PLS results showed that instrumental bread firmness (using Instron) could successfully predict quality (loaf volume and height) and sensory (softness, resilience, chewability and staling) characteristics of bread. Thus, instrumental bread firmness could be replaced to sensory analysis in evaluating bread quality.