posted 10-07-2008 04:34 PM
Ralph,The Lafayette's OSS-3 algorithm has a nice tool to display the charts. You can handscore the data, and the algorithm will give you a statistical estimate regarding the test results. OSS-3 can score many kinds of polygraph tests, including 3 and 2 question single issue Zone (ZCT) exams, multi-facet MGQT exams with 2, 3, or 4 questions, and mixed-issue screening exams with 2, 3, or 4 questions. There are specialized decision policies for these exams.
The primary purpose for the chart display is to mark and exclude artifacted and uninterpretable data from the statistical analysis.
As an added benefit, when you mark data as artifacted and uninterpretable, the algorithm will capture the number and location of the artifacts, and then calculate the statistical probability that your observed artifacts occurred by random chance alone. Its a kind of CM algorithm built into the algorithm. A statistically significant difference between the observed statistic and chance (.5 or 50/50) would mean you should review your data for countermeasures or some form of intentional/strategic distortion.
OSS-3 can also use a non-parametric ANOVA (KW-test) to test whether your RQs differ significantly between themselves. Its another algorithm within the algorithm kind of thing.
r
------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)