抄録
For decades, a variety of software development technique has been practiced. However, a number of studies to objectively measure their QCD effectiveness is rather limited. Since there seems to be many mutually influencing factors for QCD effectiveness,it is not simple to single out a specific factor and measure its effectiveness. I propose a method to measure software development techniques for their influence on QCD effectiveness in terms of a controlled major element, which has one of the biggest influencing factors on QCD the ratio manufactured by an outside organization. I also propose a model with a covariance structural analysis which allows each technique a level of influence of outsourcing ratio on QCD. This model has been applied with ISBSG 1,352 data, and 14 techniques were measured, out of which 7 techniques were found significantly influencing on QCD.
