The theory of testability transformation was developed during the proposal, together with early empirical results from the first application of testability transformation to a simple incarnation of the flag problem for evolutionary testing.

The theoretical foundation of this formulation of testability transformation has proved to be very effective, acting as a foundation on which subsequent authors built. It acted as a lingua franca for collaborative work with Korel, McMinn, Binkley, Baresel and others outside of the original project, but who became involved.

The proof obligations required by Testability Transformation are different to those required in traditional transformation because the correctness concern is not couched in terms of functional correctness. Rather the correctness of a transformation is concerned with whether it preserves the sets of adequate test data for some test adequacy criterion under consideration.

Prof Mark Harman