抄録
There exist many approaches to improve QCD of software development and it is not clear whether each approach such as some usage of technical innovation is useful for quality, cost or delivery simultaneously.In order to confirm statistical significance of a series of hypotheses on the effects of factors related with a project plan the authors propose a new covariance structure model representing causal hypotheses among the three outcome measures of QCD in software developments and the three possible factors to affect QCD as technical innovation, outsourced ratio and percentage of development effort by stage.We fit the proposed model to the data from 195 actual projects in ISBSG database.
The results show that we demonstrate relation among factors by confirming quantitatively that outsourced ratio has significantly affected technical innovation. Effort reduction is significantly affected by technical innovation. Delivery time shortening is significantly associated with a outsourced ratio. And that quality improvement is significantly correlated with percentage effort of build. We demonstrate the trade-offrelation among QCD also. The quality is improved when effort reduction and delivery time shortening.
