Today we continue to discuss software testing effectiveness measures - these are  P-Measure, E-Measure and F-Measure. We had read that the E-measure helps make P-Measure more useful.Their relative values change similarly. Moreover E-measure can differentiate the detection of more than one failure while the P-Measure regards a testing strategy as good as another if they can detect at least one failure. E-Measure assumes homogeneous subdomains whereas to maximize P-Measure, only one subdomain has the failure causing inputs.
Moreover, it has been difficult to generalize such conditions for overlapping subdomains. But in this case too E-Measure is able to identify some characteristics of overlapping subdomains testing  For example, T.Y. Chen and Y. T. Yu found that the relative performance of subdomain testing to random testing depends on the aggregate of the differences in the subdomain failure rate and the overall failure rate. The differences must be weighted by the number of test cases selected from that sub-domain. In the overlapping case, the subdomain failure rates can be higher than the overall which makes the score in favor of subdomain testing as opposed to random testing.
Lastly if the E-Measure says that the subdomain testing strategy is better than the random testing then it must be the case with P-Measure also. This leads to characterizations using P-Measure. This does not mean that the same program can get the same ranking from both measures. That said, if the failure rates are small, the E-Measure serves as a good approximation to the P-Measure. More analysis with E-measure can also be helpful if the subdomain testing strategies are do-able.