Analyzing Stability of Estimates at Completion
TABLE 6. RESULTS OF HYPOTHESIS TESTS FOR EQUAL DISTRIBUTION LOCATIONS AND CONSTANT VARIANCE BETWEEN INDIVIDUAL METRICS Hypothesis Test Stability Point (%) p-value Wilcoxon Rank Sum 10 0.9888 Brown-Forsythe 10 0.9800 Levene 10 0.9732 Wilcoxon Rank Sum 5 0.9961 Brown-Forsythe 5 0.9993 Levene 5 0.9973 Sensitivity Analysis The novelty of this study is that contract effort duration is a factor for inclusion in the dataset. The results presented thus far pertain to development efforts at least 5 years in duration. To assess the robustness of the results with respect to duration, development efforts that are 4 years and 3 years in duration are analyzed. From the list of contract efforts excluded solely due to duration, a total of eight were added for this sensitivity analysis: four contract efforts with a duration of 4 years and the other four with a duration of 3 years. Figures 3 and 4 compare the results for efforts at least 4 years in duration and at least 3 years in duration to the original results. While some differences are apparent in the earlier stages of the project, the results appear to converge to the prior results towards the end of the project. The addition of contracts that are 4 years and 3 years in duration do not significantly affect the conclusions. Stability is still achieved around the 70% completion point.
68
Defense ARJ , Spring 2025, Vol. 32 No. 1
Made with FlippingBook - Online Brochure Maker