Statistical Distributions of Software Bugs and Testing Management

Background: Software testing remains among the most significant managerial duties. Aims: This work is a statistical survey of software bugs that means to derive lessons in favor of IS/IT managers. Method: As first, we have analysed the timeseries of bugs of four releases of a large program; in the second stage we have examined the time-series of two large programs. Result: The six time-series exhibited very different shapes. A software tool demonstrated how the Wakeby and the Kumaraswamy functions best fit with the six samples. This pair of distributions has a special mathematical property that is the independence of the left tail from the right tail. Conclusion: These distributions justify empirical managerial measures for software testing from the theoretical perspective, hence they provide nontrivial support to IS/IT managers. In the next future we shall investigate the Wakeby and Kumaraswamy distributions in relation to the inner structure of the software programs.