JMP platforms and statistical methods

Graph Builder:
Mosaic plot

Categorical:
Contingency table chi-square analysis, multiple comparisons, confidence intervals for proportions

Text Explorer:
Analysis of unstructured text data

Objective

Use statistical inference for proportions, including confidence intervals, chi-squares analysis, and multiple comparisons, to compare consumers’ opinion of new gluten-free cake mix formulations compared to traditional wheat flour.

Problem statement

A food product company is interested in offering a gluten-free cake mix to enhance its offerings in that important market. It has developed three test mixes, each a blend of several alternative flours (e.g., almond, rice, sorghum, millet). Ideally, a cake made from these alternatives would very closely match the taste and texture of its traditional wheat flour cake mix. The company will move forward with development of one of these mixes if its research shows that at least 80% of consumers would find it as good as, or better than, its traditional mix.

It conducted an experiment in which 50 tasters compared cakes made with these three alternatives flours to cake made with the standard wheat mix. Across three separate sessions, each taster tasted a cake made with the standard wheat mix and then one made with one of the alternatives flour mixes. For each session, the taster indicated whether they thought the alternative mix tasted Not as Good, Same, or Better than the standard mix. The assignment of the alternative mixes to the sessions was chosen randomly for each taster.

If the taster thought the alternative mix was Not as Good or Better Than the standard mix, they were asked to describe why.