I agree with you entirely - that's the correct kind of test to employ. My point is that getting a positive on such a test doesn't give you grounds to believe someone has handled explosives. Out of every 1000 positives provided by such a test, approximately 1000 are false positives, give or take 1.