Journal of The Japanese Society for Quality Control
Online ISSN : 2432-1044
Print ISSN : 0386-8230
Volume 27, Issue 1
Displaying 1-14 of 14 articles from this issue
Message
Features
Research Papers
Contributed Paper
  • Kazuyuki SUZUKI, Tomoyuki SATO
    Article type: Contributed Paper
    1997 Volume 27 Issue 1 Pages 105-113
    Published: January 15, 1997
    Released on J-STAGE: January 26, 2019
    JOURNAL RESTRICTED ACCESS
    Condition monitoring maintenance is becoming more and more important as both the scale of the system and the effects of its mulfunction are becoming larger. Here, "monitor" plays important roles to give the information on the true state of the system which is not directly observable. But it does not necessarily inform the true state, e.g. false-alarm in good state and fail-to operate in bad state. Blackwell and Girshick [2] discuss the superiority of the values of information from two experiments, and give sufficient conditions that one dominates over another. Their idea can be applied for the comparison of the value of information from two monitors. This paper presents more mild conditions than those of^<[2] using the information on the order of loss function. The condition derived on this paper includes the results of Murakami^<[7] which deals with statistical decision problems in case of 2-states and 2-actions.
    Download PDF (746K)
Technical Note
  • Hitoshi FUJI, Akira OHMORI, Shoko OGOSHI, Ayatomo KANNO
    Article type: Technical Note
    1997 Volume 27 Issue 1 Pages 114-122
    Published: January 15, 1997
    Released on J-STAGE: January 26, 2019
    JOURNAL RESTRICTED ACCESS
    As an indicator of software productivity, the ratio of software size to software development effort is often used. The productivity evaluation with this indicator depends on the physical volume of software, does not depend on its quality. Therefore, in some cases, we get a productivity evaluation of which we cannot be convinced immediately. This paper proposes a new productivity indicator that, as elements, explicitly has not only the physical volume of software but also software quality in the broad sense. Furthermore, we make trial use of the new productivity indicator to real software systems, and we examine its usefulness. The results of this study can be summarized as follows: ・The values of the new indicator are smaller than the values of the old indicator. ・In some cases, the orders of evaluation of the old indicator and the new indicator are reversed. ・The new indicator is more consistent with developer's actual feelings than the old indicator.
    Download PDF (1072K)
Quality Report
feedback
Top