References in periodicals archive ?
One project (number 8) had reused code, and this was accounted for in the KSLOC data by counting a modified line as equivalent to a new line.
The main input for SLIM is KSLOC, whose regression with actual man-months produced an R.sup.2 of only 48.5 percent, so the SLIM model seems to be adding information.
Albrecht provides another model, showing the relationship for his data between Cobol KSLOC and man-months.
The equivalent equation for the ABC data is the following: ABC man-months = -65 + 1.47 (Cobol KSLOC) (2.99) R.sup.2 / 41.9%
Another validation that was performed was a check of Albrecht's claim of Function Points as a predictor of KSLOC. Albrecht provides the following model for estimating Cobol SLOC from Function Points: IBM Cobol KSLOC = -6.5 + 0.12 (Function Points) (6.55) R.sup.2 / 71.2%
The estimated relationship between the actual KSLOC and the estimates is the following: ABC Cobol KSLOC = 6 + 1.68 (Function Points estimated KSLOC) (4.69) R.sup.2 / 65.6%
The results for the ABC were as follows: ABC Cobol KSLOC = -5.0 + 0.20 (Function Points) (4.69) R.sup.2 / 65.6% ABC Cobol KSLOC = -13.2 + 0.207 (Function Counts) (5.85) 75.1%
The results were the following: Actual man-months = 54.1 + 0.084 (SLIM) (10.17) R.sup.2 / 92.8% Actual man-months = 11.4 + 0.195 (COCOMO-Basic) (10.82) 93.6% Actual man-months = 51 + 0.17 (COCOMO-Intermediate) (12.24) 94.9% Actual man-months = 40.3 + 0.181 (COCOMO-Detailed) (11.62) 94.4% Actual man-months = -203 + 0.451 (Function Points) (4.63) 71.8% Actual man-months = -175 + 0.426 (Function Count) (4.09) 66.2% Actual man-months = -94 + 1.67 (KDSI) (2.90) 44.5% Actual man-months = -102 + 1.68 (KSLOC) (2.83) 43.2%
Only KDSI and KSLOC did slightly worse, and these are not inputs to ESTIMACS.
In particular, readers should note the sensitivity of the regression models to project number 3, which, although almost exactly equal to project number 5 in its KSLOC count, is several times larger in its actual man-months figure.
The results for Function Points and Function Counts decreased to the 54-60 percent range, and the results for KSLOC and KDSI increased to the 56-59 percent range.
Acronyms browser ?
Full browser ?