Measuring the effectiveness of a test case
J. C. Huang · 2002
This work is concerned with the effectiveness of test cases used in program testing. It is well known that no matter which test-case selection method is used in program testing, some programming errors can still escape detection because the program may produce fortuitously correct results. One reason why a program may produce a fortuitously correct result is that it contains expressions of the form exp/sub 1/ op exp/sub 2/, and the test case used causes exp/sub 1/ to assume a special value such that exp/sub 1/ op exp/sub 2/=exp/sub 1/ regardless of the value of exp/sub 2/. In that event, if there is an error in exp/sub 2/, it will never be reflected in the test result.