Defect detection is “confirmed defects detected” in defined period of time.
Defect discovery(detection) rate is equal to no. of defects found in testing effort (Tester total hrs spent)
Defect Density = No. of bugs / Developer effort (not tester)
Bug Density = Defects in Releases / Total Story Points
(Where “Story Points” are complexity points, Nebulous Units of Time, (NUTs)), or Pair Hours.
**Defect Removal Effectiveness = Defects Removed (at the step)/ Defects existing on step entry )+ defects injected during the step x 100%
Industry Average: “about 15 – 50 errors per 1000 lines of delivered code” (refer: The book “Code Complete” by Steve McConnell)
Industry Derived Measures from ISBSG [http://www.isbsg.org]
500–8500 FP/FTE, median 1600
10–1200 KLOC/FTE, median 300
2–60 defects /1000FP, median 9.3, mean 16
.01–.95 defects/KLOC, median .02, mean .18
Median 15 hours/defect Proportions
2% extreme, 13% major, 85% minor