When virtual experiment data is obtained, it is possible to analyze the data for each individual. However, typically, analysis is performed by grouping the data into strata based on the attributes of the respondents, and using average values. If there are no significant differences in the results of the stratified analysis, treating the whole data as a whole will result in more accurate results, since there is a larger amount of data. However, if there are clear differences in the analysis results for each strata, unless stratification is performed, it will be impossible to make a correct analysis, and there is a risk of obtaining incorrect information.
In this study, we will discuss stratified analysis of virtual experimental data using optimal designs in JMP. At the time of the presentation, we plan to use examples to demonstrate the operation of custom designs for Design of Experiments (DOE) in JMP 17.
