Modeling preference of commercial toasted white corn tortilla chips using proportional odds models

摘要

Eleven commercial toasted white corn tortilla chip products from the United States were evaluated by a group of 80 consumers of age 18-35 and by a trained sensory panel. Proportional odds models in conjunction with principal components were used for internal and external preference modeling of tortilla chip consumer overall acceptance. The internal preference modeling showed that flavor was the most important attribute to consumer overall acceptance followed by the interaction of appearance by flavor and texture. The external preference modeling showed that one flavor attribute (salt aftertaste) and one texture attribute (crispness) contributed significantly to increase consumer overall acceptance, whereas one appearance attribute (instrumental color a*) significantly lowered consumer overall acceptance. The information reported in this study is important to the tortilla chip industry to produce tortilla chips with greater consumer acceptability. This study implies that proportional odds model using principal components is an alternative tool for consumer preference modeling.