Level: Beginner
【Presenter】
Shinji Fukushima, Takumi Information Technology Co., Ltd.
[Co-presenter]
Koichi Sato, CEO, Takumi Information Technology Co., Ltd.
Takuya Sato, Takumi Information Technology Co., Ltd.
Atsushi Maeda, Takumi Information Technology Co., Ltd.
Kazuhiro Saito, Takumi Information Technology Co., Ltd.
[Announcement summary]
In non-clinical studies in the pharmaceutical industry, hypothesis testing often proves the effects of multiple drug doses, and multiple comparisons are essential to tightly control type 1 errors. In particular, the Dunnett's multiple comparison method is widely used to compare the control group with multiple drug dose groups, but the detection power is detected in a test in which a large number of groups are set to compare the control group and each dose at the same time. There is a problem that is reduced.
The Williams multiple comparison method is applied when the drug effect is dose-dependent (monotonic), and since the sequential test is performed from the high-dose group in a closed procedure, it is almost detectable even when compared with the t-test of the two groups. It is an attractive multiple comparison method that does not decrease, but unfortunately it is not installed in JMP at this time.
Therefore, using JSL, a sequential test was performed from the high-dose group based on the Williams method in a closed procedure, and the calculated statistic was compared with the statistical value table for Williams multiple comparison in the book to output the presence or absence of a significant difference. We are developing an add-in. It also plans to install a non-parametric version of the Shirley-Williams multiple comparison method. The announcement will introduce the features of the add-in, including a prototype demo.
[Presenter Profile]
A system engineer who has been in charge of pharmaceutical-related nonclinical business since last year at Takumi Information Technology Co., Ltd. Engaged in development of statistical analysis software for non-clinical departments and lecturer / consulting for statistical seminars.
Engaged in in vitro and in vivo pharmacological studies in the pharmacology department of a pharmaceutical company for many years, and experienced manufacturing and marketing approval applications. In 2007, when I transferred to the research management department of a group company and was in charge of non-clinical statistical analysis, I was taught JMP by Dr. Toshiro Haga, Dr. Yukio Takahashi, etc. Recently, he is mainly in charge of development using JMP's scripting language (JSL), and plans to make a series of non-clinical add-ins that extend JMP functions.
