Some Progress in Software Testing Technology

Estimating total software defects using capture-recapture models Used by software engineers since the 1970s to estimate the number of defects in software code, the capture-recapture method was first proposed by Laplace in 1786 for biological populations. The model can be described by the following scenario. Suppose a ranger wishes to estimate the number of wolves in his park. He captures, tags, and releases as many wolves as possible during a fixed period of time called a “capture event”. After the event the wolves are released and they redistribute throughout the park. The ranger then conducts a second capture event. Suppose that n1 and n2 wolves are captured during the first and second events respectively, and m of those wolves are common to both capture events (as determined by the tags). Then the total number of wolves, N, can be estimated using the Lincoln-Peterson Estimator [1],