Xref: utzoo rec.ham-radio:3549 sci.med:3684 sci.electronics:1817 sci.misc:675 sci.physics:2724
Path: utzoo!utgpu!water!watmath!clyde!rutgers!cmcl2!brl-adm!brl-smoke!gwyn
From: gwyn@brl-smoke.ARPA (Doug Gwyn )
Newsgroups: rec.ham-radio,sci.med,sci.electronics,sci.misc,sci.physics
Subject: Re: Cancer and Electromagnetic Radiation (Long)
Keywords: Cancer, Radio, Electromagnetic Radiation, ANSI, Chemical Hazards
Message-ID: <6956@brl-smoke.ARPA>
Date: 6 Jan 88 08:35:12 GMT
References: <1077@kodak.UUCP>
Reply-To: gwyn@brl.arpa (Doug Gwyn (VLD/VMB) <gwyn>)
Organization: Ballistic Research Lab (BRL), APG, MD.
Lines: 38

In article <1077@kodak.UUCP> ornitz@kodak.UUCP (barry ornitz) writes:
>     Cause                     Expected                  Actual
>     Leukemia                     29                       36
>     Lymphatic etc. Cancers       72                       89
>     Prostate Cancer              67.6 (!)                 78
>I am not sure about the statistical differences between these numbers,

That's not too hard.  In lieu of further information, note that these are
"rare event counts".  Generally you can assume Poisson statistics in such
cases.  This means the probable error of the estimate is the square root
of the estimated count, or approximately 6, 9, and 8 respectively.
Roughly 2/3 of the time, the Actual should fall within this range of the
Expected, if there is no significant correclation.  Although the observed
deviations are somewhat larger than expected, no one measurement has
statistically significant deviation from chance at the 95% probability
level.  The composite probability for the set of three measurements may
be such that it can be taken as an indication of a correlation (I don't
want to bother with the arithmetic), but I can tell by inspection that it
won't be a really strong indication.

>"amateur radio operator licensees  in  Washington
>state  and California have significant excess mortality due to acute myloid
>leukemia,    multiple  myeloma  and  perhaps  certain  types  of  malignant
>lymphoma."

This is simply a misleading statement.  Each is significant only at
probability levels well below general statistical practice (90%, 95%,
or 99%).

>I  believe  that  other causality should be  investigated  by  the  medical
>profession before Dr.  Milham's conclusions are accepted.

Quite right.  Even a statistically significant correlation does not
necessarily imply a direct causal relationship.  Conclusions such as
"amateur radio radiation causes cancer" cannot be reliably drawn from
this evidence.  I don't know whether it does or not, frankly, and the
evidence is inconclusive.  At best, it may guide further research (has
anyone tried subjecting lab mice in large numbers to higher dosages?).