Ledi Trutna, Analytical Training Consultant, SAS
This paper discusses the author’s execution of a mixture experiment. The reasoning behind the selection of the mixture design (ABCD design in JMP) and the method of collecting the data will be described. The analysis of the data will be demonstrated in JMP. The experiment was conducted by blending four different varietals to find the best tasting red wine. A team of people collected the data and provided their ratings. The analysis will include modeling the results as continuous responses, as well as attribute responses to determine rater agreement, and discussing the pros and cons of each approach.
