USE OF THE DISCRIMINANT FUNCTION TO PREDICT THE NUMBER OF CONSUMERS WHO DISCRIMINATE |
| |
| Authors: | F M SCRIVEN M F PETTY |
| |
| Institution: | Department of Food Science and Technology University of New South Wales, P.O. Box 1, Kensington, Australia, 2033.;Faculty of Professional Studies P.O. Box 1, Kensington, Australia, 2033. |
| |
| Abstract: | Three brands of chocolate milk were evaluated for liking, sweetness, creaminess and natural flavor by 48 consumers, using 100mm line scale. The mean sweetness score for one brand was significantly different to the other two and all other attributes were similar. Discriminant analysis on scores from the line scales showed that 63.8% of the evaluations of the least sweet brand (Brand A) were classified by the discriminant function as belonging to the least sweet brand. 19.1% of the scores for Brand B and 17.0% for Brand C were classified as identifying Brand A. A simple count of the number of consumers ranking each brand as least sweet showed that 63.8% chose Brand A, 19.1% Brand B and 17.0% Brand C. This suggests that the discriminant function provides a measure of the proportion of assessors who can tell the difference or, a measure of the degree of confusion between samples. |
| |
| Keywords: | |
|
|
