Conference hydra::dejavu

Today's Globe did a report on ESP (Study finds evidence of ESP
phenomenon).  Psychologist Daryl J Bem, Cornell University, was
a coauthor of the study.  The probability that the results could
have occured by chance is less than 1 in a billion.

330 tests were preformed with a hit rate of 32% (25% would be
considered chance).  They found that creative people (students
from the Juilliard School of Music were part of the study) had
higher hit rates.  The student averaged 50% hit rates and the
musicians from the group scored a 75% hit rate.

A sender sat in one room focusing on photograph, art reproductions,
or video segments taht had been randomly selected by a computer.
The receiver sat in another room with blinders and earphones 
blocking any sound and had a steady "white noise."

"The set up was designed to isolate the receiver from any ordinary
sights, sounds or sensations, on the theory that ESP impressions
may be so faint they are easily swamped by ordinary sights and 
sounds.  Also, the images used for the experiement were chosen
to be more interesting than the simple geometrical symbols used 
in many earlier ESP experiements.  Subjects in those experiments 
often become bored and did not do well."

"Later, teh receiver was shown four or more different images, in-
cluding the one the sender had been looking at and asked to pick
theone that most closely matched what they experienced during the
test period." The reciever also sat in the room and described what-
ever popped into his/her head.

This study showed the largest effect they knew of in any experiment
on ESP.  Donald B. Rubin, chairman of the department of statistics
at Harvard, said "it was strong evidence that the effect is real,"
assuming the experimental procedures were valid.

One of the things that makes the new study especially credible is 
that Bem "is one of the true agnostics in the field, who came to
it without preconceived ideas."  

Bem said the safeguards used in the experiments "rule out, for me,
all other reasonable nonpsi explanations that have been suggested."

It is thought that this study will make it a little more socially
acceptable to do such research.

    re: .1,
    
    The areas that the promising 'remote viewing' types of experiments are 
    commonly criticized by psychologists are not usually in getting 
    near-chance results once a hit is assessed, but in the assessment of a hit.
    
    One example sometimes given (from an actual study) includes interpreting a 
    stick figure and a series of unconnected rectangles as a 'hit' for someone 
    leaning against a chain link fence.  The mind recognizes the pattern
    easily, but clearly, an extremely large number of other complex 
    situations could hit just as well; a person against a row of windows,
    a person in front of a series of columns or boxes or statues or ...
    
    It's extrememely difficult to assess 
    the degree of information content being transferred.  Anyone familiar with
    Kreskin knows how frequently he was able to match pictures in his act, by 
    methods well known to conjurors.   The interpretation, done by someone
    looking for a pattern, is neccessarily biased.   There is not yet any
    such thing as an unbiased pattern matching algorithm, as far as I
    know, although they do their best in some experiments to minimize this 
    subconscious effect by various means.   
    
    To add to this, some of the best known researchers of remote viewing
    or clairvoyance phenomena in the past were _easily_ duped by 
    self-admitted conjurors and mentalists.  Not proving that the effects
    were done by conjuring means and mundane psychologial effects, but proving 
    beyond a doubt that the experimental controls were too weak to produce 
    satisfactory data for what would be to most psychologists an 
    extraordinary effect that is at least the vast majority of the time 
    achieved by demonstrably non-paranormal psychological means.
    
    So when someone says something about positive results in a particular
    experimental setup like 'this proves ESP,' and researchers
    don't have any agreement on what ESP is, we have a situation that
    one psi critic aptly called :
    
    	PPE  --  Premature Psi Ejaculation
    
    							kind regards,
    
    							todd

Todd,

you gave a general caveat that I also recognize as a strong source of bias
in studies.

Could you, however, comment on the specifics of the test described in .1?
I had the impression that they took out this bias source by having the test
persons choose out of real pictures, thereby leaving out the form/shape
interpretation.

Or does this method introduce new bias sources?
:-)
Arie

    Hi Arie,
    
    	Do you know where the original paper was published ?
    
    	If we can get a hold of it in its original form we'll
    	be in a better position to discuss the details.
    	The quotes I've seen in the news have been too vague to
    	make any meaningful evaluation.  
    
    	Btw, I'd love to see a remote viewing study done that meets the 
    	stringent requirements for publication in a mainstream psych journal.
        As far as I'm aware there hadn't been any yet.
    	If this study has succeeded in that, even this alone is a landmark
    	event, regardless of the actual interpretation of the study !
    
    							kind regards,
    
    							todd	

    I'll have move to say about this later, but some brief comments:

    1 -- This is not a "remote viewing" experiment, that is a different but
	 related technique.

    2 -- There have been many studies, remote viewing and others, which were
         clearly rigorous enough to be published in mainstream psychology
         journals -- if they had not been ESP experiments.  Some of the
         earliest remote viewing experiments had some subtle problems, most
         of which did not seem able to account for the results but you
         could not completely rule that out.  Those flaws (which would
         probably not have been perceived as flaws in less controversial
         studies) were quickly eliminated from the procedures.

    3 -- The presentation mentioned in .1 was based on a paper which has
	 been accepted for publication in Psychological Reviews, which is
	 as prestigous a journal as exists in the field of experimental
	 psychology.

					Topher
        

    Sounds like *very* exciting stuff !
    
>    1 -- This is not a "remote viewing" experiment, that is a different but
>	 related technique.
    
    I'm very interested in hearing the details when you have the chance.
    
>    2 -- There have been many studies, remote viewing and others, which were
>         clearly rigorous enough to be published in mainstream psychology
>         journals -- if they had not been ESP experiments.  Some of the
    
    	Absolutely.  That's what makes publication in PR such a landmark.  
    	It implies they've finally managed to meet the extraordinary
    	requirements for controls that are (fairly or unfairly) imposed for 
    	extraordinary experimental effects.  
    
    						kind regards,
    
    						todd

    First:  Notes like 1801.1 is why I read in this conference.
    
    Second:  I want more description on the "[The] receiver was shown
    four or more different images, including the one the sender had been
    looking at and asked to pick the one that most closely matched what
    they experienced during the test period."
    
    Was there a second human present when the receiver was given the
    four images to select among?  I worry about a `Clever Hans' effect.
    
    There is this lavish description of how the receiver is isolated during
    the postulated information exchange, and then the writer poops out!
    Aggravating.
    
    							Ann B.

    re: .7,
    Ann, a more complete copy of the Globe article was posted in ::PHILOSOPHY
    note 312.35, however it doesn't answer your questions much better.
    
    I'm hoping Topher will have more to say to the methodology.
    
    						todd

> Was there a second human present when the receiver was given the
> four images to select among?  I worry about a `Clever Hans' effect.

I had the same thought, but it's so obvious to anyone in the psychology
biz, that no study which didn't *carefully* control for it (by being
rigorously double blind) could possibly get into PR. This would be true
for studies in *any* domain of psychology research, not just psi. If
there are flaws in the experimental technique, I'm sure they'll be much
more subtle than that.

    I agree with Mike.
    I'd be *very* surprised if there were any blatant methodological
    problems.  What I heard of the Honorton-Hyman dialogue in the past seemed 
    to  be very thorough.  Whatever might have slipped through would be 
    extremely subtle.   I feel the same way about the P.E.A.R. PK studies.
    After many fits and starts, psi is finally entering the mainstream as 
    something legitimate to try to study and understand.
    
    							todd

    Watch us (humanity) take off now... 

Personally, I think you're right. But it won't be fast or easy.

    ... ever notice how the more things change, the more they stay the
    same?  :-)

    Some followup discussion on methodology for _auto_ganzfeld_ experiments
    by Honorton, from the Internet ...
    -----------------------------------------------------------------------
Article: 39294
Path: pa.dec.com!oct17.dfe.dec.com!ryn.mro4.dec.com!nntpd.lkg.dec.com!news.crl.dec.com!deccrl!caen!sol.ctr.columbia.edu!usc!cs.utexas.edu!asuvax!ncar!noao!arizona!arizona.edu!skyblu.ccit.arizona.edu!lippard
From: [email protected] (James J. Lippard)
Newsgroups: sci.skeptic
Subject: Re: "Study finds evidence of ESP phenomenon" -- article reprint
Message-ID: <[email protected]>
Date: 17 Feb 93 05:30:00 GMT
References: <[email protected]>
Distribution: world,local
Organization: University of Arizona
Lines: 27
Nntp-Posting-Host: skyblu.ccit.arizona.edu
News-Software: VAX/VMS VNEWS 1.41
 
In article <[email protected]>, [email protected] (Alex Rosen) writes...
>It also doesn't mention whether the experimenter who shows the receiver
>the 4 choices knows which is the correct one; I assume that the paper
>wouldn't have been accepted if this were the case.  32% hits from 330
 
In the automated ganzfeld tests, the sender and receiver are in separate
soundproof rooms.  A computer selects a set of four targets, then one of the
four (a video clip or still photograph) is presented to the sender six
times over a thirty-minute period while the receiver describes whatever
imagery comes to mind.  At the end of the thirty minute period, the
receiver is shown the four targets and asked to rate them for similarity
to his imagery during the previous thirty minutes.  There is no contact
between the sender, receiver, or experimenter.  (All the experimenter does
is start up the computer program.)
 
>I saw is that having the receivers describe their "mental impressions"
>serves no purpose in the experiment, and could conceivably cause the
>experimenters to act differently.
 
Describing and recording the "mental impressions" allows transcripts to
be made and ranked against the set of targets by other persons (besides
the receiver) who don't know what the selected target was.
 
Jim Lippard              [email protected]
Dept. of Philosophy      [email protected]
University of Arizona

Article: 39305
Path: pa.dec.com!decwrl!ogicse!emory!darwin.sura.net!newsserver.jvnc.net!princeton!phoenix.Princeton.EDU!rdnelson
From: [email protected] (Roger D. Nelson)
Newsgroups: sci.skeptic
Subject: Re: "Study finds evidence of ESP phenomenon" -- article reprint
Message-ID: <[email protected]>
Date: 17 Feb 93 04:01:29 GMT
Article-I.D.: Princeto.1993Feb17.040129.8069
References: <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Princeton University
Lines: 48
Originator: news@nimaster
Nntp-Posting-Host: phoenix.princeton.edu
 
In article <[email protected]> [email protected] (Alex Rosen) writes:
>This article appeared in the Feb. 15 issue of the Boston Globe, on the 
>lower front page.  The next issue of Psychological Bulletin comes out in
>March, so we'll have to wait til then for any real discussion, but the
>article is interesting.  It made me a lot more aware of how a single study
>in the mass media can (in the public's mind) negate dozens of studies in
>scientific journals.
>
>	Study Finds evidence of ESP phenomenon
 
[Article deleted.  It has been summarized in several posts by now.]
 
>---- End of Article
>
>I have a friend who says that the Juilliard results are due to
>musicians jamming together.  I'm not sure if he was kidding or not.
>
>I didn't get the impression that Bem "went to the mass media," but that
>this is just a report of a presentation at a conference in Boston.  It
>doesn't indicate whether Bem's quotes were from the presentation or from
>an interview with the paper.  
>
>It also doesn't mention whether the experimenter who shows the receiver
>the 4 choices knows which is the correct one; I assume that the paper
>wouldn't have been accepted if this were the case.  32% hits from 330
>subjects does sounds very significant.  I don't know if 50% from 20
>subjects or 75% from even less is significant.  The other major problem
>I saw is that having the receivers describe their "mental impressions"
>serves no purpose in the experiment, and could conceivably cause the
>experimenters to act differently.
 
The experiment is a complete double-blind; the experimenter does not
know the correct choice.  The subject is shown the 4 choices by automatically 
controlled video while still in the isolation chamber, and ranks them for
correlation with his/her imagery before any contact with or feedback to/from
the experimenters.  Honorton's auto-ganzfeld experiment was good before the
"collaboration" with Hyman, and presumably better after, since they both tried
for years to conceive of flaws that had to be eliminated lest they invalidate 
the results.  
 
The description of "mental impressions" is recorded but not heard by the
experimenters.  It is heard by the "sender", and quite probably would
cause him/her to act differently.  Indeed, it was intended that this
one-way feedback would help the sender do a good job of sending.  
For details of the experimental setup, read the articles recommended by
James Lippard in this thread.
 
Roger

Article: 39307
Newsgroups: sci.skeptic
Path: pa.dec.com!nntpd2.cxo.dec.com!nntpd.lkg.dec.com!news.crl.dec.com!deccrl!caen!spool.mu.edu!howland.reston.ans.net!wupost!darwin.sura.net!newsserver.jvnc.net!princeton!phoenix.Princeton.EDU!rdnelson
From: [email protected] (Roger D. Nelson)
Subject: Re: is parapsychology bec
Message-ID: <[email protected]>
Originator: news@nimaster
Sender: [email protected] (USENET News System)
Nntp-Posting-Host: phoenix.princeton.edu
Organization: Princeton University
References: <1993Feb15.1673.20386@channel1> <[email protected]>
Distribution: sci
Date: Wed, 17 Feb 1993 04:27:28 GMT
Lines: 40
 
In article <[email protected]> [email protected] writes:
 
[regarding the Bem/Honorton paper]
 
>Assuming that the study was done OK (and I assume it was), we can say 
>that in these conditions somehing other than chance was happenning.  
>The fact that Bem, and possibly everybody else, can't think of any 
>'reasonable nonpsi explanations' does not necessarily make the 'psi' 
>explanation correct or reasonable, especially in tha absence of a
>*theoretical* explanation, and in the face of much theoretical reason
>to doubt it.
>
>When this kind of effect is found in "regular" psychology, the correct
>next step is to try to learn more by exploring the contributions of
>different aspects of the experiment (the ganzfeld, characteristics of
>the stimiuli, characteristics of the instructions, and so on), and by 
>trying to make it come and go under experimental control.  I surely
>hope this work will move beyond the manifesto and press release stage
>and take it's own findings seriously.
 
Regretably Honorton will not be pursuing this agenda, but he already 
had for a number of years been committed to learning as much as possible
of the correlates and concomitants of "successful" performance.  
For example, the standard subject or participant introduction
procedures in his lab included a "personal information form" that asked
for experiences, beliefs, and other possibly relevant data about the
subject, as well as one or more standard, widely used psychological
profiles.  He had found some correlations that bear further examination.
For a particular example of the effort to engage this issue, the subset 
of Julliard students represented a population that typically has a high 
proportion of one of the most successful psychological types. 
Some of this work has been published during the last couple of years in
the Journal of Parapsychology, some may still be in press.
 
>And, in the end, proponent of the psi hypothesis are required to
>produce a theory of psi, not just a claim that it might exists.  I'll
>start getting really excited when I am given a theory of psi with a 
>converging set of experiments consistent with it.
 
Honorton would have agreed with you.

    More Internet discussion on the auto ganzfeld methodology.
    							todd
    --------------------------------------------------------
 
In article <[email protected]>, [email protected] (Paul 
Carter) writes:
 
> In particular I'd like to know more on how the results are measured.
> Are they based on the transcripts of the receiver during the experiment,
> or on the comparison done by the receiver at the end of the experiment ?
 
On the comparison done by the receiver at the end of the experiment. In the 
scoring phase, the receiver views each of the four clips, in random order, 
then uses a joystick to position a cursor on a scale indicating similarity 
to his or her experience. This is done before the doors to the receiver's 
and "sender's" rooms are opened, and before the experimenter sees which 
clip was the target. (The program will not indicate which was the target 
until the receiver's ratings are entered.) The data are evaluated in 
terms of whether or not the target received the highest rating.
 
> How sound-proof are the rooms?
 
The receiver's room is an industrial sound-isolation room--walls, floor, 
and ceiling are made of double steel panels separated by air space. It's 
also electrically isolated.
 
The "sender's" room is separated from the receiver's room by 15-20 feet, 
and is lined (walls, floor, and ceiling) with industrial-quality 
sound-absorbing foam.
 
-------------------------------------
Ephraim Schechter
    
Article: 39468
Newsgroups: sci.skeptic
Path: pa.dec.com!decwrl!wupost!howland.reston.ans.net!zaphod.mps.ohio-state.edu!magnus.acs.ohio-state.edu!csn!boulder!ucsu!cubldr.colorado.edu!schechter_e
From: [email protected]
Subject: Re: "Study finds evidence of ESP phenomenon" -- article reprint
Message-ID: <[email protected]>
Lines: 28
Sender: [email protected] (USENET News System)
Nntp-Posting-Host: gold.colorado.edu
Organization: University of Colorado, Boulder
References: <[email protected]> <[email protected]>
Distribution: world,local
Date: Sat, 20 Feb 1993 00:21:07 GMT
Lines: 28

    I posted the abstract from a pre-publication version of the Honorton-Bem 
    report in ERIS::PHILOSOPHY as topic 312.44, and also in ::PSYCHOLOGY as
    topic 293.5, if anyone is interested.  It just mentions that what
    is reviewed is two sets of competing meta-analyses of ganzfeld data,
    and 11 new ganzfeld studies done according to the joint Honorton-Hyman
    guidelines.
    							kind regards,
    
    							todd

    Re .1
    
    >>The student averaged 50% hit rates and the musicians from the group
    >>scored a 75% hit rate.
    
    Doesn't surprise me at all.  I've been a musician since I was knee-high
    to a married grasshopper, and the most *connected* experiences I have
    ever had were when I was jammin' with a band.  
    
    -- Ruth
    
    

    Dean Radin at the University of Edinburgh reported in sci.skeptic
    that his lab had a pilot replication of the telepathy results
    reported by Honorton-Bem earlier.  His article may be found in
    QUOKKA::PSYCHOLOGY, 293.25
    							todd