Structural Test Statistics for Design Failure
John H R May, Silke Kuball, Gareth M. Hughes · Bristol Research (University of Bristol) · 1998
A first step towards linking program size and program failure rate estimation is done for a basic case of software architecture using idealized model assumptions. The statistical model used contains program size as a parameter, providing a novel type of analysis of systematic failure in systems. Development of the approach to more complex cases opens up possibilities for a new kind of system analysis with significant opportunity for cross-fertilisation between statisticians and system engineers.