cpolys Member
|
posted 05-10-2010 01:02 AM
I only have a minute but the poster does present a positive message. However, I wanted to touch on a couple statements that were made in the post. There is no explanation about what type of examination has "90% accuracy." This misrepresents what we know about screening examination, which represent the majority of examinations conducted in preemployment testing, national security, and PCSOT. These accuracy estimates cannot be generalized to screening examinations and is potentially misleading. The other statement, that "Most errors in polygraph can be blamed for an inexperienced examiner" is problematic. Polygraph is a test. No test is perfect. Tests are subject to errors in measurement. Considering that screening examinations have a median accuracy of 80%, we can expect that approximately one out of five examinations will be an error. The type of error is dependent on a number of psychometric variables (ray this is your lead in...sorry no time) Assuming or expecting that competence or experience can eliminate these errors is not realistic...but of course examiners can contribute to errors... Maybe Ray or Barry can follow up my post...no more time. Marty [This message has been edited by cpolys (edited 05-10-2010).] IP: Logged |