Tools › Project Management › Cost › Cost Estimating › Example
NASA Cost Estimating Handbook v4 Tables C-6 and C-7: power model fitted to unit costs
| Tool | Cost Estimating |
| Source | National Aeronautics and Space Administration (2015). NASA Cost Estimating Handbook, Version 4.0. NASA Headquarters, Washington, DC. Appendix C, Tables C-6 and C-7, pp. C-28 to C-30. |
| Level | Undergraduate |
Problem
Units 1, 2, 4, 8 and 16 of a production run cost 100, 80, 64, 51 and 41. Fit the power model y = A x^B by log-log regression, state the learning-curve slope it implies, and report the fitted values and the sum of squared errors in unit space.
Inputs to Enter
- Learning curve, slope source: fit from unit, cost data
- Curve data (loaded by default): 1 100 / 2 80 / 4 64 / 8 51 / 16 41
Answers
- A = 99.9830 (handbook sums with 4-decimal logs; 99.9805 unrounded), B = −0.3222, slope 2^B = 80.0 %
- Unit-space SSE 0.0348 (Table C-7), fitted value at unit 8 = 51.16; log-space R² 0.99997
- Prediction at unit 32 = A × 32^B = 32.75
Notes
The curve-data box holds this example by default; switch the Slope Source select to fitting from unit data. The Wright (cumulative-average) fit is withheld because the units are not consecutive.