Software Testing Comparison and Selection Strategy

Abstract

In the process of software testing, testing effectiveness problem of random testing and partition testing has always been the focus of the debate, in order to help the tester choose the right testing method depended on specific situation, first of all, measurement method based on P - measure and E - measure are given, and discussion analysis has carried on according to different test allocations, the result of the discussion shows that when the testing cases allocations are proportional to the size of software input domains or proportional to the software sub domains failture rate, the ability of partition testing in finding failure is better than random testing. And then, in view of the software reliability testing, measurement method based on reliability is given. Finally, composite the three measurement methods and reach a decision, the ability of partition testing in finding failure is stronger, but random testing can first find failures with large occurrence probability, and so delivered Reliability tends to be more superior.