Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!cbatt!ihnp4!inuxc!pur-ee!uiucdcs!uiucdcsp!silber
From: silber@uiucdcsp.cs.uiuc.edu
Newsgroups: net.followup
Subject: Re: Discussion of Drug Testing
Message-ID: <7300005@uiucdcsp>
Date: Sun, 5-Oct-86 04:07:00 EDT
Article-I.D.: uiucdcsp.7300005
Posted: Sun Oct  5 04:07:00 1986
Date-Received: Thu, 9-Oct-86 00:49:30 EDT
References: <119@akgud.UUCP>
Lines: 17
Nf-ID: #R:akgud.UUCP:119:uiucdcsp:7300005:000:1160
Nf-From: uiucdcsp.cs.uiuc.edu!silber    Oct  5 03:07:00 1986


Dear Abby ran a letter on drug testing a few weeks ago.  Someone wrote
in saying that only people on drugs needed to worry, and Abby royaly
flamed him.  Now for an original observation.  It may be "worse" in some
ways to have a very accurate test.  Suppose 10% of the people being tested
were on drugs, and the test had a false positive rate of 5%.  Then, about
1/3 of the people who tested positive were really clean.  As long as
everyone realized this was going on, they would have retests using more
accurate tests etc., and hold off taking action until they were convinced.
On the other hand, if the test had only 0.01% false positives, only one
out of a thousand people who tested positive would be drug free.  In this
case, the assumption might be made that, since the test was so accurate,
there was no need for a retest, especially if the person did not protest
the results.  (I am in favor of requiring the administrator of a drug test
to inform the victim as to the results, otherwise, someone could be barred
from a job for a false positive without them knowing it.)  This last
scenario appears to have occured several times in the armed services.