Level: Beginner
【Presenter】
Masa Kawasaki, Associate Professor, Oberlin University School of Business Management
[Presentation summary]
When virtual experiment data is obtained, it is possible to analyze each individual's data. However, normally, analysis is performed using average values for each stratum based on the attributes of the respondents. If there are no major differences in the stratified analysis results, treating the entire data at once will yield more accurate results since there is more data. However, if there are clear differences in the analysis results for each stratum, correct consideration cannot be made without stratification, and there is a risk of obtaining incorrect information.
In this study, we discuss stratified analysis of virtual experimental data using optimal design using JMP. At the time of the presentation, we plan to use case studies to demonstrate how to operate a custom design of experiment (DOE) using JMP17.
[Presenter profile]
After working at a marketing research and consulting company and as an organizational human resource consultant, he is currently an associate professor at Oberlin University's School of Business Management. Representative of FREELY LLC. At university, in addition to classes on business research and data science, he was also responsible for predicting students' GPA (grades) using JMP, and conducting exercises (human flow data analysis, AR application development) using the Ministry of Education, Culture, Sports, Science and Technology DX subsidy. Since his graduate school days, he has continued to conduct methodological research on the themes of analysis and design based on questionnaire surveys and questionnaire experiments, as well as case studies in which the methods are applied primarily to the fields of education and organizational human resources. https://gproweb1.obirin.ac.jp/obuhp/KgApp?resId=S000489
[Co-presenter]
Institute of Statistical Mathematics Takenori Takahashi
*Paper only exhibition
