[EMAIL PROTECTED] writes:
Hi Jackie,
I am very interested in the response of Lykken to the charge that he would
not submit his data for analysis to those opposing his position and that his
survey was fraudulent. I believe completely that people should have the
opportunity to answer charges.
I snipped considerably below for brevity only.
>Jackie Fellows <[EMAIL PROTECTED]> writes:
>Supreme Court (don't know the decision outcome--happened last
>year--maybe someone else knows). U.S. vrs Scheffer--this is a military
>case--Two amicus briefs were introduced--one supporting the view that
>the lie detector should be admissible; the second brief outlining why it
>shouldn't. Sounds very similiar to an expert witness situation to me.
The decision hasn't been handed down yet. I made a stupid blunder.
>And vested interest plays a role. The brief submitted to support the
>use was by the Polygraph Association. The informatiion can be accurate,
>but it naturally will be evidence that supports their view.
Who would dispute that?
>When one examines the sociology of science, one understands that scientists
>provide objective evidence, but they still have a bias in designing the
>study. The bias shows up in the questions asked, the topic the
>scientist chooses to study, etc. Scientists are not valuefree
>personally. People in the scientific community recognize this. That is
>why the debates, even in the sciences that you post have a higher
>status--you should listen to physicists debate the so-called "laws of
>nature." But back to the case.
Try to believe me, Jackie, I have had the displeasure of adjudicating
scientific debates on subjects I knew very little about as a technical contract
manager. It is not at all that uncommon unfortunately. I am quite well
aware of the ability of proponents to present a good face on a weak case.
>Now you stated that combined with other evidence, the polygraph is the
>most effective way of determining the truth.
@#$%@#%$
I never said any such thing.
All a polygraph can hope to do is estimate the probablility that a subject
is trying to deceive or not on a very specific question. That is not
remotely what you claim.
>Here we have three cases
>(recent ones) where there is contradiction. Of course, all three must
>be innocent, as the polygraph is the most effective way of determining
>the truth.
I repeat you are simply not making even remotely a factual argument.
Strawmen are always weak.
>In case 1, the argument will probably be that the drug test detected
>the use of cough medication or something--sorry my book is at school
>that identifies false positives for methamp. Maybe Sue knows.
Drug tests can test the wrong samples. In one case they found a subject had
used morpine because he ate a bagel with poppy seeds. Drug tests are no
more infallible than polygraphs. No tests are.
>Case 3--Louise must be innocent of any wrong doing in the death of the
>Eppen (sp?) baby.
In fact there is a great deal of evidence that Louise Woodward is innocent
of what she was charged with. She should not have been acquitted because a
polygraph supported her story but because evidence of a prior skull
fracture's healing should have lead reasonable people to find considerable
doubt. Her "confession" was the main question in the polygraph. It was
called into great question by the polygraph. A subsidiary question was
naturally whether she was the proximate cause of the Matthew Eappen's death.
That is more reliably answered by evidence from the autopsy. Jurors chose
to ignore the defense experts because they determined they were a pack of
liars. IMO they presented a very compelling case.
>And yes, the polygraph is used by government more and more--that does
>not make it any more accurate, just because government agencies or
>lawyers use them. Sounds like the "Westinghouse Seal of Approval" type
>thinking to me.
Polygraphs are used because they have shown a capability to detect lies.
Ouija boards are not used because they have not shown a similar capability.
>LOL--obstructionists like Lykken and Iacono. What are they
>obstructing--the wholesale use of an instrument that has 90% accuracy
>among the guilty subjects and a 40 to 50 % accuracy among innocent
>subjects.
I dispute your statistics but that is beside the point.
Lykken has been particularly visible in his condemnation of polygraph
reliability just as Dr. Gerdes is in the reliability of DNA in forensics.
They are best before unsophisticated audiences. I am sure most people will
remember Dr. Gerdes from the OJ trial. Both men have impeccable credentials
and both are quite challengeable on the matter of honesty. They both demand
a purity not available in the real world. I find it not unreasonable to
label such people as obstructionists. Both polygraphs and DNA have great
use in forensics. Neither can find episemological truth.
>I guess I wouldn't call that obstructing,
That's what makes for horse races.
> would call that
>safeguarding the wholesale adoption of a "quick fix" because it is
>scientific technology and therefore should be used, no matter the margin
>of error. The probability is not a high enough standard for me. That
>means the instrument is right 9 out of 10 times if you are guilty and
>only 5 to 6 times if you are innocent. (Russian Roulette for the
>innocent, IMO). Scientists, I believe, in running experiments (both
>natural and social scientists) have a higher standard than that (.05 or
>.01 in the medical field comes to mind).
In fact many tests have a lower standard of reliability than polygraphs.
Pap smears, chest x-rays, electo-cardiograms, etc. Life and death decisions
are made on the basis of such tests. You would ban them because they are
insufficiently reliable? Or should reliability be upgraded as best we can?
>Now, if someone want to settle
>for a .10 or higher probability of having an accurate test, he/she can
>be my guest. I guess I would prefer something a little more precise
>than that.
>
>But of course anyone who doesn't support a particular view could be
>called an obstructionist--I prefer the term healthy skepticism myself.
Skepticism is very healthy. Demanding the banning of an instrument because
it does not fit one's view of purity is something else again.
>BTW--the majority of who support Honts and Raskin? Opponents for the
>polygraph? I have contacted Dr. Iacono and requested information on the
>stuff you have sent.
Thank you. I will be very interested in his answer. I look forward to it.
I have not bought a rope. Yet. :-}
>But, I can respond at face value to what you sent to
>question the Lykken and Iacono survey without reading the complete
>studies of Honts and Raskin vrs Lykken and Iacono. (Sorry the stuff you
>sent came all in symbols, no text). First of all the cover
>letter--writing a letter spelling out the context in which to answer
>questions is I do not believe a methodological flaw in a study. The
>majority of cover letters in survey research provide a context in which
>to answer the questions to explain what information they want to know
>and why. One would have to see the cover letter to determine further.
>
>You answer questions (all people do) from a frame of mind that tends to
>be situation specific. It sounds like they simply asked the scientists
>to answer the questions within the framework of the legal system.
>Scientists would naturally answer differently if asked the accuracy of
>the polygraph in conducting an experimental test vrs a field test.
>Also, the polygraph used mainly in the cj field I believe is the Keller
>polygraph, which is only one form of the polygraph. This may have been
>in the cover letter and why they set the context.
>
>Second, random sampling is not the only way scientists sample a
>population. One would have to read the section on methodology to
>determine what sampling procedure they used and why they used it. Were
>they replicating the Honts survey or not?? Did they claim it was a
>random sample? If not, no fraudulent findings. In developing the
>sampling procedure they may have eliminated all well-known experts on
>both sides of the debate and only sent the survey to those scientists
>who have not focused entirely on the polygraph so eliminating Honts and
>Raskin would be reasonable. This would be explained in the study, I
>would imagine. I have sent for the study from the APA and may find the
>answers when and if I receive it.
>
>Cut-off distribution points are arbitrary points usually so the only
>problem would be in comparison which you mentioned. However, what was
>the standard deviation?? What was the basis of the cut-off points if
>not std?? Footnotes do not give me that information.
>
>Here I am making an assumption. I imagine that the Lykken study was
>conducted as a rebuttal to Honts original study and perhaps pointed out
>flaws in Honts study. The reply back to identify flaws in Lykken's
>study is SOP procedure. That is what makes science so interesting. It
>is just like any other study. There will also be those that do not
>support a study.
>
>As to why they had not released their data for reanalysis, I do not
>know. But, they are well respected in the field for other studies not
>just the polygraph and I tend to believe that they must be ethical in
>their work or they would not still be active in the field. I could be
>wrong.
>
>jackief
>--
>In the sociology room the children learn
>that even dreams are colored by your perspective
>
>I toss and turn all night. Theresa Burns, "The Sociology Room"
>
>
>
>Subscribe/Unsubscribe, email: [EMAIL PROTECTED]
>In the body of the message enter: subscribe/unsubscribe law-issues
>
>
Best, Terry
"Lawyer - one trained to circumvent the law" - The Devil's Dictionary
Subscribe/Unsubscribe, email: [EMAIL PROTECTED]
In the body of the message enter: subscribe/unsubscribe law-issues
