Frestimate accuracy

The accuracy of the prediction models depends on three things:

Input variable	Accuracy depends on
defect density prediction	Accuracy and completion of survey responses Inherent accuracy of model chosen Models with more inputs are generally more accurate than models with fewer inputs as long as the inputs are known The degree to which the survey respondents understand the survey questions
growth rate prediction	Generally the least sensitive of the input parameters
size prediction	This is not predicted by Frestimate.  The software group determines this by comparing the size prediction to the actual size once coding is complete.  This parameter is usually the most sensitive to error.

If the surveys are completed accurately the below is an indication of the typical relative error with the defect density prediction

Model/Module	Number of survey questions	Relative error when predicting escaped defect density
Basic Historical model	None.  Historical failure data must be collected from your own organization.	If data is recent, complete and similar in application type this is generally the most accurate predictor
Basic SEI CMMi model	One question	If level <2 - 83% if level >= 2 19%
Basic Industry type	One question	154%
Shortcut model	15 questions	See below chart
Full-scale model	Between 60 and 120 questions depending on how you answer initial questions	See below chart
Rome Labs model	From 44 to several hundred depending on how many surveys you answer	Depends on selection of application type factor and number of questions answered
GUESSING	None	Guessing during the early phases of development generally results in a relative error of 400% or more.  Guessing gets better as code becomes more complete.