BETH CLARKSON, Flame Thrower

Beth Clarkson is a very courteous, educated woman from Kansas who teaches math at WSU. I have exchanged emails with her for several weeks now, trying to explain to her the many ways in which self-deception can take hold and refuse to release its grip.
She has given much thought and consideration to the question of whether or not to apply for the Challenge, and we have shared a valuable discourse as a result.

This morning I received her application and claim letter, as follows:

=============================================== Claim to be demonstrated: TELEKINESIS

The applicant will demonstrate control of a candle flame by directing the heat of the flame towards a specific target. The results of the attempted control, even when successful, are subtle. Success occurs on less than half of attempts. Success or failure will not be obvious or immediately apparent to an observer in a single trial. Instead, many trials will need to be done and a database of results built up. Controls must be used in every trial to make comparisons.

The probability of the results will be computed using a standard statistical analysis. If the first Null hypothesis can be rejected, the applicant will have succeeded. If the 2nd Null hypothesis can be rejected, the applicant will have failed. Trials will be performed until one or the other of the two hypothesis can be rejected. The specific statistical analysis procedure, the level of confidence required for rejecting the null, and a minimum number of trials are among the details of the testing protocol that must be agreed to in advance by both the applicant and JREF.

-Beth Clarkson
============================================

Dear Ms. Clarkson, I am in receipt of your duly executed Challenge application and claim letter. The first problem we need to address is your proposed rate of success. I regret to say that if success occurs on "less than half the attempts", this would not qualify you for the million dollars. A success rate equal to or approximate to CHANCE would provide no proof whatsoever of the vailidy of your claim.

Hence, JREF accepts your claim, but we must reject your test protocol proposal as inadequate. Unless you can demonstrate a MUCH higher success rate during testing, your claim cannot be validated. What are your thoughts on this matter?

Sincerely, KRAMER (JREF)

a previous email

One of the issues that remained a difficult one for Ms. Clarkson was the Challenge rule which states that applicants may NOT be tested anonymously, and that all data may be made public by the JREF.

The following email illustrates her misgivings.
==============================================

Dear Mr. Kramer,
I have decided to pursue your challenge. I mailed in the application form this morning. I expect my partners in this endeavor to do so in the near future. I wanted to talk with you about the request to keep our identities private. I understand that this is your competition, and that you get to make the rules. I can also understand not wanting to be contractually bound to keep anything private.

However, I am hoping I can persuade you to keep our identities undisclosed voluntarily. It is not something that you are in any way obliged to do, obviously, but I am going to lay out the reasons why I wish you to do so.

Basically, of the three of us, I am the main one that is pushing for
privacy. None of us are professional psychics nor are any of us interested in pursuing such a career. However, I am the only professional whose career is likely to be jeopardized by such an attempt. I currently teach mathematics at a university and am pursuing a Ph.D. in statistics. For me, whether we win or lose the challenge, public revelation of my involvement in such an activity is likely to be detrimental to my professional future.

For me, it is a lose/lose proposition. Either we will show that we have no psychic abilities and I will be branded a fool professionally. Or we will show that we do possess such abilities, in which case we will be subject to huge amounts of media attention, disrupting my life and affecting the ability of others to judge my professional work on the merits of my work
alone.

In addition, my pursuit of experimentation in this area has strained the bonds of my marriage. No mean feat in a relationship that has lasted over 25 years. My husband is an adamant skeptic and revolted by my interest in psychic matters. While I have been honest with him about my activities, I am trying to keep a low profile of the activity in order to maintain
harmony at home. I even addressed the return envelope to my office rather than home.

So why am I pursuing this at all? Because, after decades of ignoring certain personal experiences, I have come to a point in my life where I feel compelled to find out for certain. I cannot explain certain things that have happened to me nor can I deny them any longer. I have to know. And in order to know, I have been conducting tests by myself for months. Every set of tests I have devised has come up with positive results.

If psychic abilities exist, then I possess them. If I possess them, I can also figure out how to measure them. I am qualified and competent to design and analyze such tests. I have conducted sufficient tests to eliminate to my own satisfaction all explanations other than the experimenter effect. While I have tried to be as honest as possible, unconscious experimenter bias could indeed explain my results.

The one thing that I cannot do by myself is conduct a blinded experiment on myself. That requires another individual. Someone who is capable of taking measurements consistently and without bias. (Incidentally, I don't feel that my partners are capable of doing this. We are working together to provide moral and emotional support to one another, but I am the only scientifically trained person in the group.)

Further, because my results indicate that the effect I am producing is subtle (perhaps 30 degrees difference on average when successful) and somewhat irregular (I estimate that I am successful in only 30% to 50% of trials) this is going to require a lengthy commitment by another individual to conduct adequate blinded trials to prove or disprove the hypothesis of telekinesis. Blinding, by the way, will not be easy. I am no actress and
have serious doubts about my ability to fool an observer regarding whether I am making a true effort or not.

Your challenge offers me more than the opportunity to collect the prize money. It offers me something worth the risk of embarrassing myself publicly and professionally by confessing a willingness to believe that such things might be real. It offers me the assistance I need to conduct scientifically valid tests of the caliber necessary for me attain certainty.

And that, in the end, is what makes it worthwhile for me to pursue this. Whether we win or lose your challenge, at the end of it I will have sufficient data to accept one hypothesis or the other.

Still, I'd prefer to keep it quiet. I find it acutely embarrassing and
somewhat shameful to admit to others my interest (and possible talent) in matters of this nature. I hope that you will consider keeping my identity private for the course of these trials. I feel a bit like a homosexual begging someone not to out him, but I'd really prefer to remain a closet psychic.

Thank you,
Beth Clarkson

another email exchange...

>Yes, one person alone can NOT, by definition, conduct a double-blind test. Can you ask a friend to assist you? A university student, perhaps?
++++++++++++++++++++++
Short answer: No. Don't think this hasn't occurred to me! Unfortunately, I have no one I feel I can rely on in this matter. I had originally hoped my partners might be able to do this, unfortunately they are simply not careful enough to entrust with the measurements.
======================
>I just want to try to get you to do some tests before submitting to the potential embarrassment of a JREF test.
++++++++++++++++++++++
I understand and appreciate your concern. Thanks. However, as I said before, I've simply come to the point where I need assistance and I'm willing to risk the potential embarrassment in exchange for that help. I hope that you are open to all possibilities, and that you do not have a closed mind.
=======================
>I am open to all possibilities. I simply see those possibilities as bleak, based upon my own limited experience in this field, and upon Randi's nearly 50 years of seeking and finding nothing. So, my mind is open, but I'm seeking something that can be verified. Science isn't perfect, but it's all we have as our sole tool for investigating weird things. It's our "candle in the dark".
I'm also entirely unconcerned with HOW your ability works, so the fact that you yourself do not understand it makes no difference to us here at JREF. No explanations are required. Only a demonstration is needed. Please do not feel compelled to offer reasons or hypothesis. All you need to do is devise a scientific test that can verify your claim.
+++++++++++++++++++++++
Thanks. Your initial 'assurances' that I am mistaken did concern me. I can understand seeing the probability as bleak.
========================
>I also know full well that you are NOT trying to fool yourself.
Self-delusion doesn't work like that, and even the smartest people on earth can be totally wrong. In fact, the very reason that so many scientists and "geniuses" are fooled is that they think they are so smart that they can NEVER be fooled. How could an Israeli conman named Uri Geller fool a dozen Stanford scientisists and researchers? He did it by knowing that their belief in their own infallibility was the key to their downfall. And the Stanford scientists did NOT enlist the participation of experts in deception, or magic, which would have helped them to see that deception was at work, and that nothing paranormal was taking place. It is VERY easy to fool yourself. So, should you fail, simply look upon it as a learning experience.
++++++++++++++++++++++++
Yes, I'm quite aware of this. As I said, it's the only explanation that I can't rule out by myself. I can be and am consciously careful, but I can't deny that I may unconsciously be making subtle choices that bias the ultimate conclusion. However, I don't assume that I am mistaken either. I have subjective perceptions that, well, that are the reason we're having this conversation. One hates to reject one's own personal experiences as being invalid without strong evidence that one is, indeed, mistaken.
If I fail, I'll at least gain that out of the attempt.
==========================
>If you're serious about becoming an applicant, Beth, the first thing to do is APPLY. One application per claim, one name per application. Three people cannot apply together for the same claim. One claim per person.
++++++++++++++++++++++++++
I understand. I mailed in the application form for myself last Saturday. You should receive it sometime this week. My partners may mail in similar claims in the near future, but I've told them it's not a rush. We'll need time to work out the details of the testing protocol first.

Beth Clarkson

Further Clarification....

The most recent email exchange between myself and Ms. Clarkson clairifies her protocol proposal...further proof that even the most vibrant, complex intellects can be subverted, under the right circumstances. This woman is no dummy. In order to help prevent people from falling into such intellectual trappings, we must somehow come to understand the process by which the believer BEGINS to believe. To understand this process is the key that will open the door to educating people in the hope of reversing this sorry, all too human tendency to believe the craziest things.
============================================

>> Dear Ms. Clarkson,
>> I am in receipt of your duly executed Challenge application
>> and claim letter.

Thanks for letting me know. I'm glad to hear that you have recieved it.

>>The first problem we need to address is your proposed rate of >>success. I regret to say that if success occurs on "less than >>half the attempts", this would not qualify you for the million >>dollars. A success rate equal to or approximate to CHANCE >>would provide no proof whatsoever of the vailidy of your claim.

On first reading of your response, I was insulted. (Do you take me for a statistical novice?). On reflection, I realized the error was mine. I obviously did not make myself clear in the two paragraph description. My apologies.

I am not expecting you to accept a claim approximate to that of chance. On the contrary, it is my profession to make the distinction between what is random chance and what is not. What I was trying to communicate is that what I consider a "success" occurs on only about 1/3 to 1/2 of my attempts, thus multiple trials are necessary to establish my claim.

A success should always lead to a positive result. Random chance will lead to a positive result 50% of the time and a negative result 50% of time. If I am "successful" on one third of my trials, I can expect to see (+) results on 2/3 of all trials (1/3 from actual successes and 1/3 from random chance) and (-) results on 1/3 of all trials. This is a testable deviation from the results of random chance alone.

An an example, consider my last 16 trials from my most recent testing protocal. I have 11 (+) results and 5 (-) results. The probability of this occurring by random chance alone is 0.105. On the other hand, it is perfectly consistent with a probability of 2/3 positive results, 1/3 negative.

If I combine the trials from my last two sets of tests (the previous protocol differs very little from the current protocol), I have 36 tests with 27 positive results and 9 negative. The probability of getting those results by random chance alone is less than 0.002.

Does this explanation suffice, or do I need to give a more detailed statistics lecture <grin>?

>>Hence, JREF accepts your claim, but we must reject your test >>protocol proposal as inadequate. Unless you can demonstrate >>a MUCH higher success rate during testing, your claim cannot
>>be validated. What are your thoughts on this matter?

I think I need to send you a detailed explanation of how I set up my tests and what results can be expected. In addition, I think you will need to see the actual experimental equipment I am using. I'm pressed for time today, but will try to box the stuff up for you this week-end.

Beth Clarkson

Detailed Protocol Suggestion...

Ms. Clarkson just sent me this, from a friend of hers who tried to help her set up a protocol...
==============================================

This is an experiment to test someone's ability to perform telekinesis. I've met some people who claim to have psychic abilities and I devised this test to check out the claim. Specifically, the subject tries to control the direction of a candle flame.

The set up is this:

I make flat rings of candle wax with a center hole that is 1.25 inches in diameter. I place a tealight candle centered inside a clear glass tumbler. The glass has straight sides that slope slightly outward. The outside diameter of the wax ring is barely larger than the diameter of the glass, so the wax ring just fits over the top, which centers the hole nicely.

I run controls as well as tests, measuring the direction of the flame by setting a target at 0° and marking the place on the ring where the wax first begins to melt (this is fairly easy and consistent to determine). I then measure the angle between the target and the first melt.

I've done a reasonable number of controls and for this set-up, the direction of the flame is fairly consistent for each candle with the first melt occurring in an approximately normal distribution around the mean for the set-up. The mean does vary with each experimental set-up, but the standard deviation is reasonably consistent and approximately 26.

The glass with the lit candle is positioned such that the flame is approximately at the eye level of the subject. The subject attempts, when a test ring is in place, to direct the candle flame towards the target. Once the experiment starts, the only touching of the glass occurs when the wax rings are changed out and I do this myself. I use 4 rings for the experiment and 4 for the control. I mark the position of the first melt and measure the angle myself, as carefully as I can.

I've run a number of variations on this experiment, trying to eliminate other possible sources that might influence the flame/wax melt. The last experiment I ran was in my house at my dining room table. I had the air conditioner off and the windows shut, thus no influence from a fan or breeze. The experimental trials were randomized with the controls, so there was no time effect. The only opening to the flame is the center hole in the wax, and the subject never raised hands above the top of the glass while the wax rings were in place. The subject is not deliberately attempting to alter the heat of the air from the flame with their breath. I don't think that would be possible anyway with this setup. The candle flame is completely contained and the melting occurs on the underside of the ring.

Anyway, the results I got from this last experiment were as follows:

Control: 77, 67, 83, 105

Test: 10, 25, 19, 38,

Running a t-test, whether with equal or unequal variance assumption, gives a p-value of less than 0.001. I'm somewhat reluctant to accept the hypothesis of telekinesis occurring in my dining room, but I'm out of ideas regarding what could be happening. Any suggestions on how to improve my experimental procedure? Or ideas about what might actually be occurring?

Thanks, Beth
========================================

I wrote back to say that this protocol may alleviate some of my concerns.

More from Ms. Clarkson...

Kramer,
Thanks for your email. I appreciate hearing from you.

I'm anxious to hear what you think. I've been thinking about the
vibration concern. Since I'm randomizing the sequence, the only real issue is that I am, in some subconscious way, vibrating the floor differently when I attempt to direct the flame than I am when I don't and that is somehow affecting the flame.

However, I ran an experiment last week-end with very interesting results wherein I was making a conscious effort not to tap the floor or anything like that. Also, because I'm only getting results about 1 time in 3, I don't think that explanation is terribly likely.

Sorry about not getting any written materials to you yet. I ended up spending all my spare time this week-end working on a particularly thorny statistics homework assignment.

This stuff is terribly difficult to measure and replicate. In addition, any sort of testing is an anxiety producing situation and anxiety or stress is likely to negatively impact success. I've been giving a lot of thought to why everyone who has previously attempted it has failed during the actual test. Mental state is paramount. That's one reason I brought in my partners. They provide me with moral support and encouragement.

Also, along these lines, I need to let you know that when we start the actual testing, I will need to move deeper into a "belief" state. There are great many things in life where one must believe something is possible before one can succeed at it. This is one of them. One crucial ingredient in successful trials is how well I can convince myself that it works beforehand. It has not been easy for me to balance between having the belief necessary to do this and the suspension of belief necessary to design good experiments.

On the other hand, it has been invaluable to be able to test my experimental designs myself.. If I were just acting as designer, I would have given up in frustration long ago. In fact, I had originally contacted Randi about testing my teacher in these things. However, he was just too hard to get to sit down and do tests. Because a large number of tests are necessary, I started designing and doing my own. That's one of reasons I starting working with the candle flame and telekinesis. Most of the other "talents" require more than one person to conduct a test.
- Beth
==============================================

One crucial ingredient in successful trials is how well I can convince myself that it works beforehand.

This one sentence speaks volumes as to the mechanics of BELIEF
and how people begin down the soggy path of self-delusion.

More from Beth Clarkson...

Kramer: Is "moral support and confidence" really needed if PSI truly exists?

Beth: If you don't understand how it works, how can you say what is and isn't necessary for success? We're talking about something that is a result of mental effort. Why would you think that the mental state of the person attempting it isn't important?

Kramer: If PSI exists, it should work regardless of how people "feel"?

Beth: I disagree, but respect your opinion. Think about this -
If I made the same statement about my efforts to win a Ph.D. in mathematics, would you doubt that my abilities in mathematics were real because I needed the support and encouragement of others in order to accomplish that feat? I assure you, such moral support and encouragement is crucial for both endeavors. I would never have gotten as far as I have in my studies without the encouragement and support of my family, friends, teachers, and mentors.

Kramer: If you're doing your work correctly, based upon proven, repeatable phenomenon, you don't need any friends, family or support group in order for it to "work".

Beth: One crucial ingredient in successful trials is how well I can
> convince myself that it works beforehand.

Kramer: Beth, read this sentence again and THINK. What does this tell us? It tells us that you have convinced yourself that this works before even looking at the results. It doesn't stand up to scientific scrutiny. What you BELIEVE will not constitute a successful demonstration.

Beth: You're right, what I BELIEVE will not constitute a successful demonstration. But what I believe will affect whether or not I can successfully acccomplish just about anything. While believing that I can do something doesn't guarantee success, if I believe that something is NOT possible, I will NOT be successful. That is why belief is a crucial ingrediant for success - not just for for what we are doing here, but in virtually every aspect of life.

Kramer: I'm anxious to see your claim tested, Beth, and even more anxious to see your response, should you fail to demonstrate your claim. I wonder if scientific proof that your claim in not valid will affect your confidence in any way. Most applicants are never swayed from their beliefs. But, YOU are an intelligent woman. I'm hoping that, should you fail, you will have come away having learned something, and that you will not chose to simply head straight back to the drawing board and re-apply for the Challenge next year.

Beth's Test Protocol Proposal

My head began to spin when I received this. I knew Randi would find it WAY too complicated, and he did. His response will be in my next posting.

-------------------------------------------------------------------------------

Experimental Protocol

Materials needed:
· Tealight candle
· Glass (clear Libby’s 16 oz tumbler with straight sides)
· Small Tealight Candle holder that fits inside glass. This helps to center the candle and raises it closer to the top of the glass
· Wax Rings – These are made from 1 ½ to 2 tablespoons of Yaley’s premium glass fill wax using mini angle food cake pans as a mold.
· Peel off letters – these are used to identify the wax rings and mark the targets on the glass.
· Seam ripper (any sharp needle-like object should do). This is used to mark the targets on the wax rings and identify the point of first melt by putting a small hole in the wax ring at that location.

To run an experiment:

1. The glass is cleaned and polished. It should be as clear as possible. A target is marked at random on the upper edge of the glass. The candle holder is placed upside down in the bottom of the glass to provide a stable centered platform to place the tealight candle on. The tealight candle flame is centered in the glass. The glass containing the candle is placed on a surface such that the tealight candle is eyelevel with the subject attempting to control the flame. The target is oriented to the subject’s preference.

2. 10 wax rings are identified with the first two being initial controls and 3 through 10 for the experimental trials. A target is marked on each wax ring using a diameter of the ring to connect the inner circumference (target for flame) and the outer circumference (target on glass).

3. If the first two rings indicate that the set up is acceptable, the subject will attempt telekinesis with the flame on 4 of the 8 experimental wax rings. The order of test vs control trials will be randomly selected from among 6 different possible balanced randomized orders that are designed to equalize any uncontrolled effects between the test and the control trials. A single die is rolled to select which of the 6 different possible orders will be used. (A table of randomized sequences and corresponding die numbers is shown below.) Prior to beginning, the subject will give the observer the sequence of the test vs. control rings that will be attempted, but the observer will not open the envelop or file until after all measurements have been completed.

4. One at a time, when subject indicates readiness, each ring is placed over the rim of the glass, centering the center hole and lining up the target on the wax ring with the target on the glass. The location of the first drop of melted wax to fall is marked by putting a hole through the ring with the seam ripper. If it is obvious to all observers and the subject that one particular drop will fall first (often only one large drop forms), then it may be marked prior to actually falling. The ordering of the wax rings (1st, 2nd, 3rd, etc.) , the placement of the wax ring on the glass, and the marking of the first melt and removing the ring should be all done by an observer ‘blinded’ to the sequencing of the test and control rings.

Table of randomized sequences.

#1 #2 #3 #4 #5 #6
1 Control Control Control Test Test Test
2 Test Control Test Test Control Control
3 Control Test Test Control Test Control
4 Test Test Control Control Control Test
5 Test Test Test Control Control Control
6 Control Test Control Control Test Test
7 Test Control Control Test Control Test
8 Control Control Test Test Test Control

After the wax rings have been marked, the angle between the target and the first melt will be measured. A mean will be computed for the test rings and the control rings and the difference between the means 0c - 0t is the test statistic that will be analyzed.

The results of these trials will be used to test both of the following sets of hypotheses:

1. Null: mc - mt £ 0° (Telekinesis has NOT occurred)
Alternate: mc - mt > 0° (Telekinesis has occurred)

2. Null: mc - mt ³ 5° (Telekinesis has occurred)
Alternate:mc - mt < 5° (Telekinesis has NOT occurred)

Some things that need to be decided on:

1. Type of analysis: There are two possible analysis procedures that I consider appropriate .
a. Binomial distribution: Consider only positive versus negative results. That is, if 0c - 0t > 0, that is considered a “success”, if 0c - 0t < 0 that is considered a “failure” . We can then test the null hypothesis of set 1 with p0 = .5 and the null hypothesis of set 2 with p0 = .5278 . Standard binomial computation of probability can be used. This can be set up in EXCEL as follows:
Hypothesis set 1: =BINOMDIST(# successes, # attempts, .5, 1)
Hypothesis set 2: =BINOMDIST(# failures, # attempts, .4722, 1)

This test has the advantage of being very robust and the only assumptions required are independence of the trials and a constant probability of success, both of which are appropriate. It has the disadvantage of not taking into account how large a difference occurs between the means of the control rings and the test rings and it will require a large number of trials to be conducted before any conclusion can be reached.

b. Paired t-test: This test has the advantage of requiring fewer trials to come to a conclusion because it will take into account the actual values. The assumptions required – independence of the trials and the measurements have an approximate normal distribution are both appropriate. This can also be set up in EXCEL.

2. Absolute value of measurements vs. -90° to 270°. There are 360° in a circle. The target is set at 0°. The measurements can occur anywhere on the circumference of the circle. It isn’t appropriate to use 0 to 360 as the measurements, because measurements that occur close to the target, but on opposite sides would then be recorded as being very far apart rather than close together.
There are two possible ways to deal with this potential problem. One is to record the measurements in negative as well as positive degrees. This leads to another potential problem – i.e. measurements opposite the target that are close together are recorded as being very far apart. I have found that using a range of -90° to 270° and assigning the positive value to the side with the majority of measurements to resolve this problem works well as long as only 3 quarters of the circle need be measured. This stills leaves a potential problem in experiments where measurements in all 4 quarters of the circle occur, but that seems to be quite rare. Still, it is a difficult problem when it does occur.
The other potential way to deal with this is to record all measurements as absolute values (i.e 0° to 180° only). This works quite nicely, but will reduce the variance of the measurements within an experiment. This wouldn’t be a problem for either of the analysis procedures I proposed above.

3. The sequence of trials (controls and tests) need to be unknown to the observer and anyone who makes measurements. It should also be determined in advance. I recommend that the subject determine the sequence and give the sequence to the observer in a sealed envelope or a password protected computer file prior to each experiment. The envelope or file would not be opened until after the experiment is complete and all measurement are complete and have been accepted.
Note: It is quite possible that the observer will be able to distinguish the control from the test trials during the experiment.

4. Probability level necessary to reject null hypotheses. Each set of hypotheses needs a level of confidence and it’s associated a to determine when the null hypothesis can be rejected. Do you want to be 99% sure you are making the correct decision? Or 99.9%? Or 99.99%? The higher the level of confidence required, the higher the number of trials that will need to be conducted.

5. Number of trials to be conducted. A minimum and maximum number of trials to be conducted should be established prior to the series of experiments. This will need to be based on the type of analysis used and the a-level set for the hypotheses.

=============================================\

Ooof.

Randi wrote the following to this applicant in response:

++++++++++++++++++++++++++++++++++++++++

No. Each ring should be marked with 12 radii -- numbered. Wax ring is
set in place, then a position from 1 to 12 is chosen by random means.

The trial is done.

Independent "double-blind" persons decide -- only viewing each wax ring
-- which of 1 to 12 sectors was affected.

James Randi.

Beth's reply to Randi's comments -

No. Each ring should be marked with 12 radii -- numbered. Wax
ring is set in place, then a position from 1 to 12 is chosen by random means.
==============================================
BETH: Big problem. I'm not claiming to be able to affect the natural variation of the flame to that large an extent. This is subtle effect.
+++++++++++++++++++++++++++++++++++++++++++++
RANDI: The trial is done. Independent "double-blind" persons decide -- only viewing each wax ring which of 1 to 12 sectors was affected.
==============================================
BETH: I Don't have a problem with independent "blinded" observers to establish where the melting occurring. Shucks, that's the main reason I'm applying for this! A "double-blind" experiment would require that I be "blinded" as to which wax rings I'm trying to effect. Not possible. The best we can do is a single-blind experiment. Double-blind experiments are typically used in medical trials of medications where a placebo can be used (sometimes even with similar side-effects to make it less obviously a placebo) and neither the subject nor their evaluator knows who's getting the placebo and who's getting the real thing. Hence the term "double-blind".

I'm not claiming a "power" sufficient to convince anyone, even me, in a single trial. I thought you guys wanted scientific evidence! Scientific evidence requires repeated trials, careful measurements and statistical analysis to establish whether a seemingly subtle difference between groups is real or just random chance. So what's the objection to using those in an experiment? - Beth

JREF"s latest protocol suggestion...

We're trying so hard to accomodate this woman, but it seems she wants to see if CHANCE goes her way, although she may not realize that's what she's going for. Self-delusion never came in so bright a package before.

Here's my latest email to Beth Clarkson:
==============================================

OK Beth,
The numerous tests you propose are deemed entirely unnecessary in proving your claim. If you insist upon using such a complicated, convoluted and unnecessary protocol in a series of
tests, I must ask you, on behalf of James Randi and the JREF...
How do you intend to pay for such trials (which would obviously entail hundreds of hours and require the hiring of observers and
experts over a long period of time)?
I direct you to rule #6 of the Challenge:
All expenses such as transportation, accommodation, materials,
assistants, and/or all other costs for any persons or procedures incurred in pursuit of the reward, are the sole responsibility of the applicant. Neither the JREF nor JR will bear any of the costs.

I strongly fear that unless you can agree to a MUCH simpler protocol, your claim cannot be tested. The conditions you are insisting upon are far too complicated. How about a simple, binary test? Our proposal: we videotape a test in which you are given a to-be-determined number of opportunities to move the flame either LEFT or RIGHT. Simple. The video recorder will be
placed ABOVE the candle flame, facing down so that it will be quite clear which direction, if any, the flame leans towards.

Your claim was that you can, by using your mind, affect the movement of a candle flame. Fine. Only one test is required, as follows:

A candle is arranged to be present at testing. You light the candle when you are ready to demonstrate your claim. We turn on the video recorder.

You then tell us in which direction you will cause the flame to move.

Then, when you're good and ready, DO IT.

-Kramer, JREF

Re: JREF's latest protocol suggestion...

KRAMER said:

We're trying so hard to accomodate this woman, but it seems she wants to see if CHANCE goes her way, although she may not realize that's what she's going for. Self-delusion never came in so bright a package before.

Here's my latest email to Beth Clarkson:
==============================================

OK Beth,
The numerous tests you propose are deemed entirely unnecessary in proving your claim. If you insist upon using such a complicated, convoluted and unnecessary protocol in a series of
tests, I must ask you, on behalf of James Randi and the JREF...
How do you intend to pay for such trials (which would obviously entail hundreds of hours and require the hiring of observers and
experts over a long period of time)?
I direct you to rule #6 of the Challenge:
All expenses such as transportation, accommodation, materials,
assistants, and/or all other costs for any persons or procedures incurred in pursuit of the reward, are the sole responsibility of the applicant. Neither the JREF nor JR will bear any of the costs.

I strongly fear that unless you can agree to a MUCH simpler protocol, your claim cannot be tested. The conditions you are insisting upon are far too complicated. How about a simple, binary test? Our proposal: we videotape a test in which you are given a to-be-determined number of opportunities to move the flame either LEFT or RIGHT. Simple. The video recorder will be
placed ABOVE the candle flame, facing down so that it will be quite clear which direction, if any, the flame leans towards.

Your claim was that you can, by using your mind, affect the movement of a candle flame. Fine. Only one test is required, as follows:

A candle is arranged to be present at testing. You light the candle when you are ready to demonstrate your claim. We turn on the video recorder.

You then tell us in which direction you will cause the flame to move.

Then, when you're good and ready, DO IT.

-Kramer, JREF

Click to expand...

It's been confirmed:

Well, it's been confirmed: I'm losing my mind.

This just in from Beth Clarkson:

========================================
Mr. Kramer,

My proposed protocol is a starting point for discussion. It has evolved over many months and I expect it will continue to do so. In fact, I told my partners to expect it take from 6 to 12 weeks of discussion with you to finalize the protocol we will use. That includes time for you to consult with experts of your own choosing in regard to the experimental design.

What I am interested in is a fair. honest, and accurate assessment of my claim and I am not adverse to changing the protocol in any way that will lead to an improvement in that direction. How complicated or convoluted the tests are is not a main concern for me. Any aspect deemed unnecessary
by both of us can certainly be discarded.

The world is a very complicated place and seemingly unconnected events may, in fact, be tied together. The protocol I have proposed is an attempt to isolate the flame and any effect I may be having upon it from the rest of the world, be it the flapping of a butterfly's wing in Brazil or my own bated breath as I await the fall of a drop of wax. The convolutions that you disdain as unnecessary and too complicated are the result of my best
attempts, with limited funds for equipment, to achieve such isolation. If you have suggestions on how to simplify the experiment while maintaining that isolation of effects and obtaining accurate objective measurements, I will be more than happy to consider them.

As far as paying for such trials, may I remind you of an email conversation we had a few weeks ago about that very subject? You assured me that I would not be required to pay for observer's time and expert's consultation. That, in fact, you had volunteers who would donate their time and services to JREF. In particular, you indicated that you could find an observer in my area willing to donate their time to this cause. If this is not the case, I will be quite disappointed. Costs for travel, expenses in regard to actually conducting the test such as wax, candles,
measurement instruments, videotape, etc., those I expect to bear. Any expenses that you need to make and expect to be reimbursed for must be cleared with me first.

I am quite aware of the scope of what I am proposing. I am glad that you are now cognizant of it as well. The numerous hours required is why I am unable to find anyone to help me with observations. I could easily find someone to observe a single trial. But a single trial is not sufficient to prove or disprove a hypothesis. However, the number of trials needed could
be reduced by a factor of three if you would reconsider allowing more than one person to participate in the same claim and combine the results of experiments with me and my partners. Another possibility to reduce the burden of performing observations is if you can find more than one observer
in my area. I don't see any reason that we must use the same observer for every trial.

Now, in regard to your proposal of having a camera pointed down at the flame and facing down so that it will be quite clear which direction the flame moves. I don't object to using a camera in this matter. There are, in fact, certain advantages to doing so. However, a flame moves not simply left or right, but in four dimensions: left/right, forward/backward, up/down, and through time. If you wish to use a camera and judge the results in this way, a considerable amount of work will be required before
experimental trials can begin.

Criteria for determining what will constitute a "success" must be
established. Are you proposing a measurement via computer analysis (you did at one point indicate you had someone willing to write such software) which can be analyzed statistically? Or simply a yes/no decision? A yes/no decision will require considerably more trials to reject one of the two null hypotheses that will signal an end to the testing. On the other hand, constructing a measurement suitable for analysis requires a good deal of time and thought in developing both the criteria for how to measure the direction of a flame.

In addition, before beginning any actual trials, tests of any such system will be necessary. I would have to devise some other way to affect the flame to the same small degree I am claiming I can and test your proposed system to see if it can, in fact, reliably distinguish such a difference from the normal variation of a flame. While all of this can certainly be done, it will require a considerable amount of effort, and I am not sure
that the results would be an improvement over what I have already developed. However, if this is your wish, I have no objection to doing so.

Let me now review some of the convolutions and complications of my current protocol and why they are in place.

Multiple Trials: The reason for this should be obvious. As I stated in my original claim, the effect is subtle and intermittent. I feel that I am able to influence a flame only about 1 time in 3 attempts and I am only able to achieve a small difference when I am successful. In order to ascertain whether such a difference is real, multiple trials are necessary.

Flame set up: A flame is very sensitive and easily influenced by a variety of movements. I have placed the flame in a glass and covered the glass with a wax ring to eliminate effects from other sources and ensure that I am not unconsciously manipulating the flame with either my breath or my hand movements. Only the small hole in the center allows outside air movement to affect the direction of the flame. Even so, I find it necessary to close the room off or unexpected breezes can ruin the experiment.

Wax rings: This is actually the part of my protocol that has undergone the most revision and the part I am most receptive to changing. I finally arrived at the wax rings as a way of bringing uniformity into the measurements of the direction of the flame. The rings are quite uniform and a close fit to the top of the glass allowing for consistent placement. I am able to center the flame with sufficiently precision beneath the wax ring to get measurements that are accurate and reliable.

(Incidentally, I've been having a student perform a second set of
measurements on the rings used in my experiments and am pleased to report that my measurements do not show any indication of bias.)

There is a small amount of subjectivity in both the placement of the rings and the identification of the point of first melt. A "blinded" observer placing the rings should eliminate any bias I may unconsciously have when I perform an experiment myself. I'm afraid that it will be obvious to an observer when I am attempting to influence the flame and when not, so bias
in identifying the point of first melt is still possible but an observer
that you provide will, at least, not be biased in the same direction that I am.

In addition, the wax rings are easy and cheap for me to produce. When I first started setting up experiments last spring, I had hoped to get thermal probes to measure the temperature directly and accurately. Unfortunately, when I looked into it, I found that the costs of such probes was prohibitive.

The use of multiple wax rings within a trial is done primarily for
statistical reasons. Sample means have a well-established distribution and knowing the distribution of the test statistic ( the difference of the mean of the control rings and the test rings) facilitates the analysis of the results.

Randomization of the rings within a trial: This is standard protocol for any type of designed experiment. It's actually a major pain for me because switching between attempting to influence the flame and not influence the flame is one of the hardest parts for me about an experiment. But it's necessary to even out the effects of any and all uncontrolled factors that might influence the results whether we are aware of those factors or not.

Now, you know the reasons for the way I have set up the test protocol the way I have. I am more than willing to listen to suggestions for improvement. I also sent you, along with protocol, a list of 5 things that need to be decided upon. While some of them are specific to the protocol I have outlined, some - such as confidence levels - will need to be decided upon regardless of the final protocol agreed to. Have you given any
thought to any of those?

Beth Clarkson

==========================================

I think the most alarming thing about this is that she has now somehow seen fit to drag one (or more, for all we know) of her students into this personal delusion of hers. My assessment of her level of intelligence has just dropped a few thousand miles.

I wonder how the student's parents would react if they knew that they sent their offspring to school to learn something, and instead, they are playing with fire.

Most alarming. Most disturbing. Indeed - most infuriating, at this point.

That's my opinion, anyway.

My REPLY:

Dear Ms. Clarkson,

JREF enlists the assistance of qualified observers who volunteer their services for ONE TEST.

JREF does not ask people to spend inordinate amounts of time (as in DAYS & WEEKS) testing claims that can be verified or refuted in a single test.

That said, I think the entire matter is now moot, based on several statements you have made in this latest email, such as:

"As I stated in my original claim, the effect is subtle and intermittent."

INTERMITTENT? What does this suggest to you, Beth? Think about it.

Here's another statement you made that requires closer observation:

"I feel that I am able to influence a flame only about 1 time in 3 attempts and I am only able to achieve a small difference when I am successful."

Beth, 1 out of 2 is CHANCE. You claim 1 out of 3. We require more proof than what you yourself purport to be able to offer, at best. You will NOT pass preliminary testing. It would be a complete and utter waste of time for all those involved to proceed when you are claiming to be able to do something that will not pass testing.

Additionally, as regards your statements about "confidence levels ", the matter of the applicant's "confidence level" is one of the most common excuses used after failing a test, and for backing OUT of testing, as well. You seem to have this "escape portal" prepared already.

If your claim cannot be verified under proper observing conditions (as per the Challenge rules), then it will have nothing to do with "confidence levels". The reasons for your failure would center solely on the fact that the movement of the flame has nothing whatsoever to do with your attempts to influence it.
The flame would act as it does if you were in China, sleeping comfortably in your own bed, or Fishing with John.

Beth, we're getting very close to rejecting this claim. We will not cater to your whims regarding what you believe to be the
necessary complexity of the protocol design. It is our concerted opinion that no such lengthy and involved tests are required.
Either you can do what you claim to be able to do, or you can't. It's that simple, and it doesn't need 1,000 tests to establish.

-Kramer, JREF

Randi's comments:

I have forwarded several of the outstanding comments offered by forum members to Randi for comment - particularly those from persons with expertise in statistical analysis who support some of what Beth is insisting will show that something other than chance is at work here.

Here is his concise comment, and, for as much as I would like to see Ms. Clarkson's claim tested, I agree wholeheartedly with Randi's viewpoint.

=============================================

Understood, but if someone is claiming to get results less than chance would call for, she has NO ability! Should we test her for having a huge failure rate….?

James Randi.

Vanishing Act

We have not heard from Beth Clarkson in several months now.

Dis? I think not -

Kramer,
Why did you post that I had vanished? I've been hanging out on the JREF forum since December, posting regularly. The last email (dated Dec 1, 2004) I had from you regarding my claim is below, which indicates that you were not interesting in testing the claim I originally made. I've emailed you since then, though not recently. I haven't managed a better protocol yet (thermal probes are expensive), and very few observed trials under my existing protocol. Why would I bother emailing you that there is nothing new yet? So what's up with dissin' me on the forum?
> Beth

==========================================

Beth,

It was no dis. I haven't heard from you in months. That's what I posted.

-Kramer, JREF

=========================================

> No, you posted that I had vanished. It implies that you've looked for me and couldn't find me. People don't usually say that someone they haven't heard from recently, but still occasionally see around has vanished. How often do you want me to contact you to say "nothing new"? I'll be happy to send you an email as often as you like, but I don't anticipate much progress before June or July, and that's assuming I'm successful in developing a better test that JREF might find acceptable AND still have significant results.

As it stands now, you've told me my current protocol is unacceptable to you and your suggestions have all been unacceptable to me.

---Beth

===========================================

You're quite right, Beth. We're nowhere near agreeing on a test protocol, and at this point, I'm pretty much convinced that we never will arrive at one. You've totally dropped the ball on the application process. There has been NO ACTIVITY or correspondence of any kind in months, and 12 months from the submission of your application your file will be closed.

You'll then have the opportunity to apply again 12 months later.

-Kramer, JREF

FILE CLOSED

Never is a long time, but I don't anticipate having the resources to upgrade my protocol any time in the near future and I don't think you and I will agree to anything I can put together with the resources I do have available. That's why I hadn't made any contact with you for the past few months.

If you want to close the file now, you may. I'll contact you if and when I feel I have something worth having you take a look at it. I would appreciate being able to keep the lines of communication open regardless of whether you wish to classify my claim as open or closed. Any chance of that?

Beth

========================================

OK, Beth, your file is now CLOSED. Thanks for being honest about everything, and good luck in your quest to decipher what is true. No hard fellings, I hope, and PLEASE re-apply if you find something more concrete that we can test at a future date.

Despite appearances, our minds remain entirely open, as do the lines of communication.