I've been a JMP user since JMP4, and I've also been using SAS since before JMP. Before retirement, I conducted statistical education using various JMP methods and promoted JMP within my company. I presented the results of this work at JMPer's meetings as appropriate. After retirement, I established a self-employed consulting company and continue my activities. Recently, I've come to realize that the ability to statistically analyze real data in many academic fields in Japan (excluding clinical statistics) is collectively declining. For me, experimental design was indispensable for learning statistical theory and for practical application. Last year's theme was related to the design of experimental sample sizes, and the challenge was to enable experimental researchers to estimate the minimum detection sensitivity MDD for a new Latin scale experiment with a fixed sample size, using the RMSE, which is the square root of the mean sum of squares of residuals in past Latin scale experimental data. The t-distribution functions available in JMP scripts include non-central t-distribution functions, so it's difficult to say that this can be done using them, but I was very pleased to find that it could be done by cleverly using the DOE's "Sample Size Explorer". This year, we will focus on atavistic partitioning experiments.
