Knowledge Partitioning and Knowledge Transfer Mechanisms in Software Testing : An Empirical Investigation