Defense Acquisition Research Journal #108

Analyzing Stability of Estimates at Completion

60% complete and EAC SCI ($) to achieve stability within 10% deviation at approximately 20% complete. Utilizing ES metrics provided stability results similar to EVM metrics. Ultimately, the graphical analysis of cost and schedule performance stability suggests no significant difference between the different metrics in estimating cost at completion for long duration development efforts. The statistical comparisons aid in reinforcing these preliminary conclusions. Comparison Analysis Table 3 summarizes the descriptive statistics of %EAC(10) and %EAC(5) for EVM and ES. Table 4 presents the results of the Wilcoxon Rank Sum test and Brown-Forsythe and Levene tests. The comparison shows no statistically significant difference between the distributions and variances of %EAC(10) and %EAC(5) for the EVM and ES methods. This suggests the two methods have comparable accuracies in predicting final schedule for long duration development efforts.

TABLE 3. SUMMARY OF MEANS, STANDARD DEVIATIONS, AND CONFIDENCE INTERVALS FOR STABILITY RESULTS OF EVM AND ES METHODS

Stability Point (%)

Standard Deviation

Lower Bound

Upper Bound

Method Metric

Mean

EVM SPI($)

10

72.168

21.360 63.718

80.618

EVM SPI($)

5

83.754

14.953

77.839

89.669

EVM SCI($)

10

70.577

19.903

62.704 78.451

EVM SCI($)

5

83.261

15.364

77.183

89.339

ES

SPI(t)

10

71.934 22.411

63.069 80.799

ES

SPI(t)

5

83.779

14.797

77.926

89.633

ES

SCI(t)

10

71.016

20.289

62.990 79.042

ES 89.255 Note. ES = Earned Schedule; EVM = Earned Value Management; SCI = Schedule Cost Index; SPI = Schedule Performance Index. SCI(t) 5 83.477 14.606 77.699

66

Defense ARJ , Spring 2025, Vol. 32 No. 1

Made with FlippingBook - Online Brochure Maker