Journal of The Japanese Society for Quality Control
Online ISSN : 2432-1044
Print ISSN : 0386-8230
Volume 52, Issue 2
Displaying 1-7 of 7 articles from this issue
Invited Article
  • Takenori TAKAHASHI
    Article type: Invited Article
    2022 Volume 52 Issue 2 Pages 84-89
    Published: April 15, 2022
    Released on J-STAGE: November 25, 2022
    JOURNAL RESTRICTED ACCESS
     Hyper design is a general-purpose, analytical design based on hyper structure functions. The feature of this method is optimization by mathematical programming using a hyper structure function, which is a dual structure function based on one or more design factors designated as super factors. Sometimes, the function to be designed is very complex. In this case, instead of approximating the points in the entire section with a single function, the following methods should be considered as alternatives. Multi-headed design: design using column directional functions for each individual level of the hyper factor. Excerpt hyper design: Design by excerpting the necessary subintervals and approximating each with an appropriate function. And the hyper design of the general form, which combines and integrates the above to fit the situation, is the following design. Combined hyper design: Design that combines multi-headed design and excerpt design
     This spaper discusses these issues.
    Download PDF (2206K)
  • Takenori TAKAHASHI
    Article type: Invited Article
    2022 Volume 52 Issue 2 Pages 90-95
    Published: April 15, 2022
    Released on J-STAGE: November 25, 2022
    JOURNAL RESTRICTED ACCESS
     This paper first discusses selective dual-seid causal analysis. Principal components are used in front-side causal analysis, and factors are used in back-side causal analysis. Together, they form a dual-side causal analysis. What is important in this analysis is the following. It is not possible to do anything about the principal components. However, although the factors are latent, they can be taken action by extracting them and then deciphering them. This analysis reveals what needs to be done. Next, how to obtain useful insights by using the data collected for selective dual-side causal analysis in an SEM-like approach is clarified. In this process, it is important to conduct causal analysis from three perspectives: worm's eye, animal's eye, and bird's eye. In this study, this type of causal analysis that approaches from multiple perspectives is called compound eye causal analysis and discuss its concept and structural application.
    Download PDF (1053K)
  • Akira OGAWA, Takenori TAKAHASHI
    Article type: Invited Article
    2022 Volume 52 Issue 2 Pages 96-101
    Published: April 15, 2022
    Released on J-STAGE: November 25, 2022
    JOURNAL RESTRICTED ACCESS
     This paper discusses the education of comprehensive design methods for design and hyper design.A comprehensive design method means a set of steps starting from error reduction, screening of design factors, modeling of functions for design, optimization by mathematical programming, checking the realization of the solution, and modification of the solution. After experientially learning these, education for the basics of hyper design is provided. In hyper design, it is necessary to deal with complex model structures that include product terms (interaction) and quadratic terms. It is shown that virtual learning material (simulator) is effective for learning this alone, in a short time, and efficiently. This education ensures that engineers understand through simulation experiments that hyper design is optimization under statistical variability.
    Download PDF (1353K)
  • Sho KAWASAKI, Takenori TAKAHASHI
    Article type: Invited Article
    2022 Volume 52 Issue 2 Pages 102-107
    Published: April 15, 2022
    Released on J-STAGE: November 25, 2022
    JOURNAL RESTRICTED ACCESS
     Analytical methods for surveys that include causal structures are approaches based on three perspectives:worm’s eye, animal’s eye, and bird’s eye. Combining multiple approaches to uncovering these causal structures and using them in different ways depending on the purpose of the research is called compound eye causal analysis.
     In this study, we first organized the analysis method based on the purpose and viewpoint of the survey. Next, we applied the following two methodologies to an online survey case study with a large number of questions involving causal structures. One is the analysis approach of worm’s eye:selective dual-sided causal analysis based on the vital few and the trivial many, and the other is the analysis approach of animal’s eye:analysis of three quasi-SEM models to developmental approach the hypothetical model of SEM. In conclusion, discussed the results of compound eye causal analysis and clarified the characteristics of these methodologies.
    Download PDF (730K)
Reportage
feedback
Top