Level: Intermediate
【Presenter】
Takenori Takahashi, Visiting Professor, Keio University
[Announcement summary]
Questionnaires to improve the current situation require both consequential and causal questions. Then, if the causal relationship between the two is grasped by regression analysis, it is possible to select the target to be dealt with. In that case, the items that you missed cannot be covered after the investigation, but if there are unnecessary items as a result, you can ignore them after the fact. For this reason, if the question items are prepared without omission in consideration of not overburdening the respondents, the result will be a large number of question items and a high correlation will appear between the items. To solve this problem, Takahashi and Kawasaki have proposed "multi-group principal component regression analysis". The essence is that the principal component is obtained for each group under a rational group configured so that the correlation within the group is high and the correlation between the groups is low, and this is an important principal component in principal component regression. Is selected, and the one with a large absolute value of the factor loading for the selected principal component is selected as the main item to be dealt with.
Sometimes this key item is dense, which can be addressed by factor analysis. The causal analysis is the front side causal analysis when the principal component is used, the back side causal analysis when the factor is used, and the two-sided causal analysis when both are combined. In this presentation, we will introduce the theory for that and a concrete method using JMP.
[Presenter Profile]
The author has been conducting research and practice of QM (quality management), SQM (statistical quality management) and design theory (joint research with many companies and management guidance, etc.) for half a century. Since the 1990s, we have proposed a new design paradigm, Hyper Design, researched its mathematical HOPE theory (Hyper Optimization for Prospective Engineering), and provided its support software, HOPE-Add-in for JMP, to SAS. We are co-developing with. From the above, a new design method is realized by the trinity of super design, which is the idea, HOPE theory, which is statistical mathematics, and HOPE-Add-in for JMP, which is a support tool.
Design is often misunderstood as a special activity by special people because of the high threshold. In order to overcome this and allow many people to master the design, the author is developing new educational methods as well as theoretical research. It is a hands-on education using real teaching materials (paper helicopter, paper glider, coin shooting, etc.) and virtual teaching materials (fly ball simulator, etc.). This educational program (from the basics of statistics to super-design) is an easy-to-understand and interesting education for performing visualization analysis and design by JMP in many situations. This education has been conducted at domestic and overseas universities (Keio University, Yale University, Tokyo University of Science, University of Tsukuba, etc.) and many companies for more than 30 years, and its effectiveness has been confirmed.
* For those who wish to have a more detailed understanding, we have prepared a dissertation. If you wish, you are the presenter Takenori Takahashi Please contact us directly.
