
336 Probability Methods for Cost Uncertainty Analysis
the preceding section. The right half of Figure 9.8 shows the empirical corre-
lations between Y and its pairings with X
1
, X
2
,and ε. They were computed
by the Excel Correl function. Observe how the empirical correlations compare
favorably with their analytically derived values, as summarized by matrix A
in Section 9.3.2. Differences between them are due to random errors endemic
in all simulations. From a practical view, such differences are negligible.
9.4 Summary
The importance of correlation as a critical consideration in cost uncertainty
analysis cannot be understated. Seen herein,