Many risks due to human factors are latent in a software development project. If suitable management can be performed to these risks and a software development process can be improved continuously, we will lead the project to improve in the productivity and quality of software product. In this paper, we analyze the process data collected from actual software development projects, and clarify the process factors which affect the quality of software product, by using multivariate analysis. Further, we also discuss a method of quantitative project evaluation based on a software reliability growth model, which helps us to give useful quantitative measures for determining project completion.

1. S. Yamada and M. Takahashi: Introduction to Software Management Model, Kyoritu-Shuppan, Tokyo(1993).
2. S. Yamada and T. Fukushima: Quality-Oriented Software Management, Morikita-Shuppan, Tokyo(2007).
3. S. Yamada: Elements of Software Reliability - Modeling Approach, Kyoritsu-Shuppan, Tokyo(2011).
4. P. B. Moranda? Prediction of software reliability during debugging, Proc. 1975 Annual Reliability and Maintainability Symp. , pp. 327--332, 1975.
5. S. Yamada and A. Kawahara: Statistical analysis of process monitoring data for software process improvement, Intern. J. Reliability, Quality and Safety Engineering, Vol. 16, No. 5, pp. 435--451(2009).
6. S. Yamada, T. Yamashita, and A. Fukuta: Product quality prediction based on software process data with development-period estimation, Intern. J. Systems Assurance Engineering and Management, Vol. 1, No. 1, pp. 72--76(2010).
7. S. Yamada and J. Yamakawa: Statistical analysis based on software process monitoring data for effective project management, J. Global Perspective on Engineering Management, Vol. 1, No. 2, pp. 44--50(2012).

Principal Component Analysis Multiple Regression Analysis Structural Equation Modeling Logistic Regression Analysis Software Reliability Growth Model Geometric Poisson Model