RE: Low levels environmental

[Kjell Johansen <kjell.johansen@wepco.com>]

You should definitely continue to use negative values in your
statistical analyses.  Not only does DOE/EH 0173T recommend it but this
recommendation goes back to August of 1980 when the HPS Committee Report
HPSR-1 (1980) "Upgrading Environmental Radiation Data" which was
released as EPA 520/1-80-012 recommended using all values as measure,
whether negative or positive.  I strongly recommend Section 6 which was
written by a group who came from various regulatory and scientific
bodies as ( EPA, NRC, NBS, DOE, ORNL, EML) as well as others (Alabama
Power Co, State of Illinois, and Idaho Falls).  If you have looked at
the latest ANSI standards in the HPS Newsletter, you will notice that
all bioassay standards also recommend reporting and using negative
numbers in statistical analyses.

I think that the push to use <MDA for negative values arises from the
perceived problem of having to explain "negative" radioactivity to the
public should the need so arise.  However, in the end, I believe that we
must insist on good statistical procedures to obtain at good science.  

I have been using negative numbers in our REMP for several years.  If I
make a frequency plot of the airborne I-131 data I get a fairly nice
bell shaped curve around zero.  Just what I'd expect based on our I-131
emissions.  Also, the averages for all of the I-131 results are not
different from zero.  If I threw away the negative numbers, I would get
a positive average.  

I believe that EPA 520/1-80-012 still is available from NITS.  I suggest
that you show Section 6 of this and all of the other relavent and
supporting standards to whomever is asking you to substitute MDAs for
negataive numbers to back up your use of negative numbers.   

Kjell Johansen, Ph.D.
kjell.johansen@wepco.com
Sr. Radiological Engineer
Wisconsin Electric Power Co.
Milwaukee, WI
************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html