Hidden Truth/Hidden Truth.iso/data/newspostings/1601-1700/1643

Internet Message Format | 1996-06-20 | 7.7 KB

Path: news.demon.co.uk!dispatch.news.demon.net!demon!usenet2.news.uk.psi.net!uknet!usenet1.news.uk.psi.net!uknet!tank.news.pipex.net!pipex!news.mathworks.com!nntp.primenet.com!news.primenet.com!lippard
From: [e-mail] (James J. Lippard)
Newsgroups: alt.paranormal,sci.skeptic,alt.paranet.psi,alt.paranet.ufo
Subject: Re: CSICOP -- Dennis Rawlings -- sTARBABY
Date: 20 Jun 1996 07:36:01 -0700
Organization: Primenet [phone]
Lines: 147
Message-ID: <4qbnkh$[e-mail]>
References: <4ok4jv$[e-mail]> <4pkatq$[e-mail]> <4pmmc2$[e-mail]> <4pnm8r$[e-mail]>
Keywords: Mars effect
X-Posted-By: [e-mail]
Xref: news.demon.co.uk alt.paranormal:19757 sci.skeptic:72928 alt.paranet.psi:5497 alt.paranet.ufo:53911

From [e-mail] Thu Jun 20 07:32:16 1996
Date: Thu, 20 Jun 1996 10:31:30 +0000
From: Suitbert Ertel <[e-mail]>
To: "James J. Lippard" <[e-mail]>
Cc: [e-mail]
Subject: Re: Jim's sci.skeptic answer


 >In article <4pkatq$[e-mail]>,
 >   [e-mail] (James J. Lippard) wrote:
 >
 >>If you didn't know there was still a controversy, then you haven't
 >>been paying attention.  Two books have been published on the
 >>subject this year (a pro book from The Urania Trust and a con book
 >>from Prometheus).  Everybody who knows what they are talking about
 >>agrees that there is evidence of selection bias in Gauquelin's
 >>work, the remaining controversy is over whether Gauquelin's bias
 >>accounts for the entire effect, what Ertel's eminence trend data
 >>means, and what the CFEPP results mean.  The believers are
 >>challenging the CFEPP results and claiming that CFEPP's data shows
 >>a Mars effect.  The skeptics are claiming that the CFEPP data put
 >>the final nail in the Mars effect's coffin.

 >>I haven't come to a conclusion myself, though I am leaning towards
 >>the "final nail" view.

             Suitbert  asked:
 >Why are you wavering?
             Jim's reply:
(1) I've been wavering for a long time.  I've never been fully persuaded
 that there is a real "Mars effect" in the data, only that the data
 hasn't been explained by anyone and merits further study.


             S:
 >Why are you leaning towards the "final nail" view?
             J:
 (2) It seems to me that Gauquelin's work doesn't really support the
 "Mars effect" on its own terms, because of the selection bias, which
 appears substantial enough to account for the entire effect.  What's
 left is your eminence trend results, and I wonder how strong that
 really is.


              S:
 >What would help you come to a final conclusion?
               J:
(3) More time to look at this stuff more carefully, as well as training
 in statistics or the evaluation of some experts. -- Jim Lippard
---------------------------------------------------------------------------

Jim,

Your answer (2) above challenges my comment:

Most data selections in ordinary reserach are biased to a certain degree.
Nevertheless most research results are valid. The reason is that
bias effects are generally small. That is, the fact of bias itself is
no sufficient reason to reject research reports. The only question
is "Is the reported effect entirely - or only partially - due to an author's biased
selections?"

I assume that Benski's and  Nienhuys' report in their book on Gauquelin's bias
made you believe that it [G's bias] "appears substantial enough to account for
the entire effect. "

Unfortunately the authors do not address substantial issues.

 - Gauquelin did not select cases for his own study. He suggested
   to add/delete cases for the CFEPP's study. The CFEPP generalized
   from Gauquelin's suggestions for improving their sample to
   Gauquelin's selections of cases for his own former studies. This
   was illegitimate, to say the least.

-  Ken and I have shown by objective citation counts that  the CFEPP's
   athletes selections were biased. Gauquelin became aware of
   the CFEPP's biased selections without citation counts. His
   planetary claim (and the protocol of 1982) required that the CFEPP
   sample should consist of truely high achievers. So he was bound to
   suggest that the CFEPP delete certain mediocre athletes that they
   had admitted and to add certain high achievers that they had neglected.
   As more high achievers than low achievers are born
   when Mars is in G-sectors - assuming the Mars effect is effective -
   his suggested additions must be expected to contain more Mars G-cases
   and his suggested deletions must be expected to contain less Mars
   G-cases. This difference re Mars percentage between suggested additions
   and deletions as such cannot be contested (except if no evidence
   for the Mars effect existed).

   Now, my impression is that Gauquelin's suggestions were guided
   - in addition, not exclusively -   by his awareness of Mars sectors
   of his suggested cases. But even this question should be carefully
   investigated before final conclusions are drawn. I think we might
   find that, say, 50% of Gauquelin's suggestions of improvement
   were justified, the rest was biased. The percentages may be different,
   but the proportion "100% Gauquelin bias, 0% CFEPP bias" - as the
   CFEPP and JWN apparently want to make readers believe,
   is clearly inconsistent with existing evidence .

  By the way, Gauquelin's bias might be understood as a desperate attempt to
   prevent the CFEPP from demolishing his life work. He had reasons to
  worry.The skeptics had excluded him from their investigation at an
  important stage: They had not shown their list of selected athletes prior to requesting
  birth dates from town halls - a clear breach of the protocol. The CFEPP's
  decisions were not only biased, not only unfair, they were illegitimate.
  Should we not also consider that Gauquelin made this attempt a couple of
  months before he committed suicide which was apparently another
  desperate attempt to solve certain problems. His committing suicide does not invalidate
  the validity of his earlier research. His exaggerated suggesting to
  improve the CFEPP's sample does not invalidate his own previous reserach either.

>  What's left is your eminence trend results, and I wonder how strong that
>  really is.

    The eminence trend is strong since

    - my 1988 discovery of a Gauquelin selection bias in his data
     was based entirely on my eminence trend test. Even the skeptics
     acknowledged the validity of this finding thus implicitly
     acknowledging the tool that I had applied.

    - my 1996 discovery of a Gauquelin selction bias in the Belgian
      CP's data is again entirely based on my eminence trend detector.
      In a correspondence with the Belgian skeptics even Dommanget
      (the Chairman of the Belgian skeptics) maintained that their
      data was clean. My eminence finding proved that he was wrong.

    -  I found significant differences of Mars G-percentages among
     low and high eminent athletes in CSICOP's U.S. athletes'
     sample whose overall Mars G-percentage was low. The same
     difference (statistically significant) was discovered in the
     CFEPP's overall low eminent sample.

     In sum, the eminence trend test was strong enough to reveal both,
     biased selections by Gauquelin as well as on  the skeptics' side.
     Note that the eminence trend test could never work if the Mars effect
     did not exist.

 Does this information help?

 Critique by any reader of this message is welcome.

 Suitbert
--
Jim Lippard    lippard@(primenet.com ediacara.org skeptic.com)
Phoenix, Arizona  http://www.primenet.com/~lippard/
PGP Fingerprint: 35 65 66 9F 71 FE 50 57  35 09 0F F6 14 D0 C6 04