Wowbagger Reviews a Pro-Parapsychology Book

[Ersby]

Oh, wow, I got 765636775. What are the odds of that? One in a billion, precisely. It's got to be signifficant. Those odds don't happen by sheer chance.

765636775 can be split into three - 765 636 775

the one in the middle is a palindrome, and on either side are two similar (only one digit difference) numbers, and the difference gives the whole number a slightly asymmetric quality that I think is very pleasing.

 

[SpuriusMaximus]

Not to mention that it contains 3 times 7, the number of God, and 3 times 6, the number of Satan, but only two 5's, the number of humans. There's a certain symbolism in having in it the full duality of good vs evil, divine vs demonic, salvation vs damnation, with an incomplete and imperfect humanity caught in the middle of it all. The only other digit is a 3, clearly a subtle hint and reminder of the trinities that are in there

Kinda just shows how much you can adlib to assign signifficance to something which, as Bob is my witless, was genuinely spit out by the first and only run of that program snippet

 

[Wowbagger]

Significance or deviation from chance, compared to _what_?

Compared to a grid where all the squares would have the same size bubbles.

That, I believe, would be the answer. Ertel seems to think one should expect almost perfectly even grids, by chance alone. (Seems like he would also expect to get 5 heads and 5 tails in 10 coin flips, but don't quote me on that.)

The funny thing is: That is exactly the trend that happens, as a subject does more trials! Only when the number of trails is small, say 180 of them, do the "significant results" come out! And only a small percentage are "more significant" than the others!

 

[SpuriusMaximus]

Well, that's saying that my 1 random number from 0 to 1 billion is signifficant if it's not exactly 500 million. And it ain't.

Seriously, that's what bugs me the most. The apparent significance of 1 data point by itself.

For whoever doesn't know that already (much as I don't expect many on JREF), when an experimental data set essentially proves "it would be too incredible a coincidence" it needs _two_ sets between which the coincidence would exist: the predicted set and the actually measured set.

 

[fls]

Here's the problem. And it's the same problem that I brought up with the ganzfeld data, Sheldrake's telephone telepathy, and a problem that Ivor touched on briefly in this thread.

The statistical analysis is performed under the assumption that the binomial test can be used to determine probability, such that improbable results are held to be statistically significant. And the assumption is made that the frequency of the numbers is equivalent to the probability that they will be drawn. However, it has never been established that these assumptions are correct. And in some cases (such as Sheldrake's telephone telepathy) it can be demonstrated that they are grossly incorrect. In order for the binomial assumptions to be valid, each number must be equally likely to be guessed (Ertel's own data shows that some numbers are more likely to be guessed than others), and each number must be equalling likely to be drawn, on each trial. Yet, we already know from the US draft that we can we woefully mistaken as to the adequacy of our mixing strategies (and the description of Ertel's experiments doesn't even sound adequate, let alone that it isn't demonstrably adequate), which means that from draw to draw, it isn't equally likely that each number could be drawn. It could be that the Just Drawn Number (JDN) is likely to be placed off-limits, slightly decreasing the probability of drawing that number. Combining that bias, with the cognitive bias of avoiding calling the JDN, would slightly increase your hit rate. I'm not saying this did happen, I'm just saying that a little thought allows you to come up with an understanding of why the binomial assumptions can not and should not be assumed.

If you look at it as a collection of people whose ways of mixing and drawing balls leads to unequal probabilities for each number, from draw to draw, one begins to form different conclusions about the "significance" of the binomial tests.

It also must be pointed out that there is another obvious interpretation to "significant" results that are obtained when the subject/recorder is unmonitored, that disappear once they are monitored.

Linda

 

[Limbo]

From Professor Suitbert Ertel:

Wowbagger: If you want mainstream science to accept this work, your best bet is to ask this: What could be a possible mechanism for these results? ….

Without a testable hypothesis, it does not matter how striking you think the results are, nor how statistically significant you measure them to be. You will always be seen as merely playing with fluctuations in the noise.

Ertel’s reply: Confirming one's observations by repeated measurements is one thing, explaining them is another matter. If mainstream scientists only acknowledged observations that already have pertinent explanations, by providing plausible mechanisms etc., then science would stagnate. Scientists who treat results without mechanisms as "merely playing with fluctuations in the noise" should not be regarded as having fertile attitudes.

Wowbagger: Remember, I am not describing my own attitude, here. I am describing the general attitude of the scientific community. The scientific community has a penchant for reliable results.

Ertel’s reply: Scientists testing parapsychological issues share this penchant. They dislike unreliable results. A certain proportion of scientific observations are unreliable, even though they are real. The only straightforward question is: Why may observations be unreliable? Why are weather predictions unreliable? Why are predictions of comet crashes on earth unreliable? Why are predictions of creative discoveries in science unreliable? Why are evolutionary changes unpredictable? All this unpredictable stuff occurs. Occurrences may be factual even though they have not been predicted and even though they are not predictable in principle despite maximum effort of the entire scientific community.

Wowbagger: If you could apply psi to the situation, someone could use "only their brain" (or wherever the source of psi's power comes from), to get most of the flips to come up heads, even after a tremendous number of trials.

Ertel’s reply: If you know x conditions giving rise to a phenomenon P which requires x+y conditions to become fully explained, you will not be able to make accurate predictions of P if you are only aware of those x conditions and unaware of those y conditions. If your brain is involved (x condition), among others, when P occurs, you cannot force P to occur by mere brain activity.

Wowbagger: Do you really think funding is an issue?

Ertel’s reply: Yes it is. The number of psi-gifted people in an unselected adult population showing statistically significant psi effects (in my standard ball test) is roughly 15%. Among these, only 10% are extremely psi-gifted (in my standard test), which means that only 1.5% of an unselected adult population shows robust repeatable psi effects. Repeatable means here that when you use the participant again and again for, say, 20 sessions, you will obtain significant deviations from chance in 18 or 19 or even in all 20 sessions. I might provide examples.

Now, you have to pay, say 100 people, for their psi test participation, $20 each, $2000 in total, in order to find, on average, one participant capable of repeating extremely significant results without much cross-situational fluctuation.

Wowbagger: If a benefactor gave you all the funding you needed, what would you do with it? What can you not accomplish with the resources you have now?

Ertel’s reply: I would test an unselected sample of 3000 people. The additional costs for experimenters and data analyzers might raise the required amount to $12,000. Only then would I continue my testing, if possible under skeptical supervision (even under JREF supervision), with 4-5 psi-gifted participants.

Ivor the Engineer: ...and most of the tests were done at home with practically no controls against cheating or bias...

Ertel’s reply: Only two of those 39 runs with form generating instructions were done at home, without experimenter control.

Ivor the Engineer: Are the bags used for the experiments pictured anywhere?

Ertel’s reply: On the following youtube video that I made roughly a decade ago an Indian psi-gifted boy has been tested by using the standard ball-drawing procedure.

http://video.google.de/videosearch?...anormal&hl=de&emb=0&client=firefox-a&start=30

I tested this boy again three years ago – now a young man, an expert in informatics –

and he still obtained a statistically significant surplus of hits.

Hans Mustermann: Ok, maybe I'm not getting something obvious, but this claim I really don't understand. ….So what was that hypothesis there? 15 of 39 looked like they weren't created by chance... compared to what? What was the formula you were fitting them against? Oh, it's aesthetic value?

Well, then that makes it even weirder. Because then a whole matrix becomes a single data point. That's really what you're measuring: the aesthetic value of the whole resulting pattern. It's not a whole data set, it's not a graph, it's just one data point: how aesthetic the whole picture was. Just like, say, the Monal Lisa is one single data point, not a whole data set.

Ertel’s reply: Hans does not seem to understand the procedure, I am afraid he needs an extra course in statistics. The chi squared procedure and its background must be comprehended, otherwise I would need much additional time and space to make myself understood.

Hans Mustermann: But, to give credit where credit is due, all the above is just a variation of Feynmann's famous quote: "You know, the most amazing thing happened to me tonight. I was coming here, on the way to the lecture, and I came in through the parking lot. And you won't believe what happened. I saw a car with the license plate ARW 357. Can you imagine? Of all the millions of license plates in the state, what was the chance that I would see that particular one tonight? Amazing!"

Ertels reply: This quote is beside the point, since deviations from chance in the ball test means deviations from precise predictions of mean chance expectation. Given 5 alternatives of equal probability, as in this experiment, the null hypothesis prediction is 20% hits (plus or minus a certain estimated chance error, the size of which will depend on the number of trials).

Hans Mustermann: Was there any analysis done on the _whole_ data set, as opposed to cherry-picking individual data points? E.g., were those 15 out of 39 pretty matrices picked by people with psi abilities, while unskilled normal people generate only, say, 5 pretty pictures out of 39?

Ertel’s reply: Due to lack of funding, I did not test psi-ungifted participants. I did test myself and I am not psi-gifted in the standard procedure. My eight form-generating runs with 480 trials each had the following results:

1 p = .03

2 p = .008

3 not significant

4 p = .o2

5 not significant

6 not significant

7 not significant

8 not significant

To explain, e.g., p = .008 , this means: 8 of 1000 random runs will produce the empirically observed hit rate above chance expectation.

Thus, my overall result may be considered as merely marginally significant. I would not dare conclusions about form-generating psi effects in my own case. But I myself am a poor psi effect producer. Individual differences are large and reliable.

 

[Limbo]

Another response from Professor Suitbert Ertel:

fls, Philosopher, wrote:

"...from draw to draw, it isn't equally likely that each number could be drawn. It could be that the Just Drawn Number (JDN) is likely to be placed off-limits, slightly decreasing the probability of drawing that number. Combining that bias, with the cognitive bias of avoiding calling the JDN, would slightly increase your hit rate."

Ertel's response:

Fls's idea is reasonable. I never heard any critic raising this concern, although a related concern has been raised based on opposite assumptions. The two assumptions are as follows:

Fls's assumption:

1. Participants avoid calling just-drawn numbers (JDN), i.e., they avoid
calling number x on trial i+1 after drawing x on trial i.

My analysis shows that this is true.

2. Participants tend to place balls with JDNs off-limits.

I tested this; again this tendency is significant, although weak.

3. Since participants avoid calling x, say number 3 which may be a JDN, they
will call 1 or 2 or 4 or 5.

Each number is written on 10 balls. Since a ball with number 3 has been placed "out of reach", the probability of drawing one of these four numbers is not 10/50 = 0.20, but 10/49 = 0.2040.

If "JDN call avoidance" and "placing JDN balls out of reach" would occur with each of 60 trials of one run, participants would obtain on average 12.24 hits instead of 12 hits. Does this matter?

My psi-gifted participants obtained up to 24 hits per run on average, average hit scores of one of my stars were in five successive test sessions 22.0, 24.2, 21.4, 21.4, 24.2, with 480 trials in each session.

Thus, fls's estimate, even with extreme presuppositions as given above (JDN calls are not always avoided, JDN balls are not always placed out of reach) an influence by fls's clever mechanism would be slight, if present at all; anyway, negligible.

Other critics' objections;

1. Participants will subconsciously place JDN balls in some position where they can reach them with above-chance probability, their subsequent juggling the balls might be insufficient.

2. Participants tend to avoid calling JDNs, but whenever they do call JDNs their subconscious memory will guide them to the place where they had put them.

I have been considering this possible source of bias from the start and would have discontinued this ping-pong ball research if I had discovered any JDN influence; even slight effects would have discouraged me with this approach. I did not find JDN effects. I tested this bias hypothesis in various ways, e.g., by assigning calls to participants. An assigned sequence of calls 1, 2, 3, 4, 5, 1, 2, 3, 4, 5 ... would be less advantageous for JDN storage and hit scores than an assigned sequence 1, 1, 1, 1, .... But no difference of hit rates was found.

 

[Limbo]

From Professor Suitbert Ertel:

I tested the young Indian, Kannan, in 1998 (my control), in 1999 (a local teacher's control), and in 2001 (my control). My standard analysis of ball test data considers JDN effects. Aside from total scores, scores from JDN occurrences and scores from non-JDN occurrences are obtained separately.

The results are as follows:

JDN | non-JDN
-------------------------------------------------------------
Calls hits proport| calls hits proport
-------------------------------------------------------------
1998 235 56 .42 | 485 207 .43

1999 432 119 .28 | 1848 500 .27

2001 132 56 .42 | 348 106 .30
-------------------------------------------------------------

You see that Kannan obtained almost the same proportions of hits (expected .20) for JDN and non-JDN calls in 1998 and 1999; only in 2001 was the proportion of hits for JDNs larger.

Kannan's calls of JDNs (disregard hits here) was larger than for student samples (32.6%); in 1999 this proportion dropped (18.0%) and rose again in 2001 (27.5%). For student samples, the percentage of JDN calls is only 13.0%. Interestingly, young people do not tend to avoid calling just drawn numbers, only at a later age the idea comes up that the person should imitate randomness which includes the wrong assumption that a number which has just occurred will less likely occur at once again.

Another instance of disproving an influence by JDN effects is the comparison of the upper half of hit scoring students with the lower half of the student sample. The JDN hit proportion of high hitters is .26, the non-JDN hit proportion is .26 (no difference). The JDN hit proportion of low hitters is .19, the non-JDN hit proportion is .18 (negligible difference). We would expect JDN effects for high hitters, larger hit scores for JDN calls compared with non-JDN calls. This difference does not show up.


In the two-handed ball-drawing experiment, the participants are instructed to regard the numbers as coordinates of a 5x5 matrix into which one dot is put down for each trial, whose location is given by the respective coordinates. The numbers drawn left define rows, the numbers drawn right define columns.

The participants are told that the dots, across 180 trials which is one run, will accumulate in the matrix cells. Dot accumulations might
be random, but the participant is told to try to obtain, by mental activity such as hope, desire, concentration, as far as possible,
non-random accumulations which eventually might form an orderly and perhaps pleasant pattern. The participant should rely on her
individual judgment of orderliness and pleasantness, no external norm need to be considered.

The instruction does not include the task to call particular rows and columns; instead, the participants are told to wait for a form or pattern (“… eventually might form an orderly and perhaps pleasant pattern”). However, the participants are not told to consciously avoid calling particular numbers. So it occurred that some participants intended to fill certain cells of the matrix, they wished to draw, say, number 1 from the left bag and number 5 from the right bag.

In order to exclude call influence I conducted some tests without feedback. This condition reduces the probability of particular mental influence on the result to almost nil. I found significant patterns even under such no-feedback conditions.

If I had time and money I would conduct such experiments with children in order to exclude such influence by mental construction altogether.

Other examples of “autonomous” psi activity are strong positive results with two bags of balls under the condition “the sum of the numbers drawn left and right should be 6”. Chance probability .20 as usual. In this case the two draw events must be “entangled”, the right hand must know what the left hand draws and/or vice versa.

 

[Limbo]

A clarification:

Wowbagger: Without a testable hypothesis, it does not matter how striking you think the results are, nor how statistically significant you measure them to be.

Ertel: My hypothesis: There is an anomalous (currently unexplained) human ability to identify numbered balls hidden in opaque bags, and a gifted individual can use this ability to build up strongly visible and statistically significant patterns in a 5x5 matrix by consecutive ball-drawing from two bags, where chance alone would yield only an insignificant random scatter.

 

[steenkh]

If such psi-talented people as this Indian guy really exist, I cannot understand why the JREF million has not been taken long ago.

Why not show him off in two simple and cheap tests conducted in India, grab the million bucks and use the money for proper scientific research?

 

[Ivor the Engineer]

Perhaps it's just my closed-minded engineering brain, but all I'm seeing in these results is evidence that some of the participants sucked at drawing numbered ping-pong balls at random out of a bag.

Thus what is required is to alter various aspects of the experiment (e.g., bags, balls and other factors) until the draw frequencies match what the statistical model predicts they should be. I.e. find out what makes the anomaly disappear. Only then can you sensibly form a hypothesis as to what was causing the anomaly in the first place.

 

[Limbo]

Passing this along for Damien:

===========

Ivor writes: "Perhaps it's just my closed-minded engineering brain, but all I'm seeing in these results is evidence that some of the participants sucked at drawing numbered ping-pong balls at random out of a bag."

This is so remarkably perverse that I must regard it as heavy-handed irony or sarcasm. On reflection, though, perhaps it can be read as a re-statement of a truism: Information = -(Entropy)

Or did Ivor really *mean* it? Rather like: "All I'm seeing in the Olympic 100 meter results is evidence that some runners suck at falling down and breaking their legs or haring off in the wrong direction with their pants pulled over their heads."

Ivor: "Thus what is required is to alter various aspects of the experiment (e.g., bags, balls and other factors) until the draw frequencies match what the statistical model predicts they should be."

Great plan! You could also test the hypothesis that sighted people can see well enough to walk an obstacle course by shutting all the lights off, until the number of accidents and concussions matches what the statistical model of total blindness predicts they should be.

 

[Wowbagger]

Limbo, I would like you to ask Suitbert Ertel a very important question, if you could:
Do you consider your work, that I commented on, to be of genuine scientific value?

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

If he answers NO:

* I will then promise to leave Ertel alone, from now on; as I would have no more interest in him, myself.

* However, I would also like you to pass on a message to Damien Broderick: That he should no longer consider Ertel's studies to be evidence that psi is ready to "come in from the cold".


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

If he answers YES:

Then, I would then like you to ask Suitbert a follow up question: What part of the scientific method do you not understand?!

Seriously, I do not enjoy sounding that rude. But, take these quotes into consideration:

Confirming one's observations by repeated measurements is one thing, explaining them is another matter. If mainstream scientists only acknowledged observations that already have pertinent explanations, by providing plausible mechanisms etc., then science would stagnate. Scientists who treat results without mechanisms as "merely playing with fluctuations in the noise" should not be regarded as having fertile attitudes.

The scientific method is supposed to help us develop new ideas that can both explain and predict things.

If someone does not have such a "new idea", then how is that attitude supposed to be more "fertile"?

My hypothesis: There is an anomalous (currently unexplained) human ability to identify numbered balls hidden in opaque bags, and a gifted individual can use this ability to build up strongly visible and statistically significant patterns in a 5x5 matrix by consecutive ball-drawing from two bags, where chance alone would yield only an insignificant random scatter.

I meant: A testable hypothesis that can boost or clarify the "anomalous ability". Something that could, in theory, be replicated in independent labs, etc.

Without that, how are we supposed to know that "psi" really is the cause behind the results? How are we supposed to rule out the following?:
* Bad experimental design (unintended bias on certain balls, for example)
* Deliberate cheating
* Paredolia or Confirmation bias
* Random chance (Hey, it could happen...)
* Erroneous recording
* An idiometer type of effect
* Etc.

In my review, I explain an alternative hypothesis: The results of psi research will only be viewed as significant by those who choose to believe in it. I outline how we can directly test and falsify this hypothesis, as well.
And, I invite Limbo, Damien, Ertel, and anyone else to comment on that.

(I call it a "theory" in the review, because it was in a section discussing theories, but the semantics of the situation are not important.)

It can be found, here: http://www.mitchlampert.net/News.aspx?R=191 , under the heading title "Theories of Psi".

 

[Ersby]

A quick note of thanks to Mr Ertel for taking the time to answer everyone's questions.

 

[fls]

Ertels reply: This quote is beside the point, since deviations from chance in the ball test means deviations from precise predictions of mean chance expectation. Given 5 alternatives of equal probability, as in this experiment, the null hypothesis prediction is 20% hits (plus or minus a certain estimated chance error, the size of which will depend on the number of trials).

This is wrong. This was my point earlier. You have mistakenly assumed that 5 alternatives gives you a chance of 20% that each ball will be drawn. However, because of mixing and other issues that are are inherent part of this experiment, each ball in that bag does not have an equal chance of being drawn, which also means that each number does not have an equal chance of being drawn. This is related to why so many people have trouble with the Monty Hall or 3-door problem.

http://en.wikipedia.org/wiki/Monty_Hall_problem

Linda

 

[fls]

Another response from Professor Suitbert Ertel:

fls, Philosopher, wrote:

"...from draw to draw, it isn't equally likely that each number could be drawn. It could be that the Just Drawn Number (JDN) is likely to be placed off-limits, slightly decreasing the probability of drawing that number. Combining that bias, with the cognitive bias of avoiding calling the JDN, would slightly increase your hit rate."

Ertel's response:

Fls's idea is reasonable. I never heard any critic raising this concern, although a related concern has been raised based on opposite assumptions.

You have missed my point. My point is not that the issue of JDN explains the relationship. My point is that inadequate mixing means that the JDN is no longer equally likely to be drawn as any other number/ball in the bag. And that this problem is not confined to the JDN, but merely represents a contribution to the larger issue of bias, such as different picking strategies, which has to be assumed to be present unless demonstrated otherwise.

All you have demonstrated with your two-hand picking study is that there is a bias present in the picking of balls. And since it takes only a little bit of effort to think of non-psi causes for that bias, it seems foolish to attribute that bias to psi.

Attempting to look for this after the fact is ultimately futile. Medicine is rife with research that demonstrates this, whereby "controlling" for biases a posteriori still reveals associations that are later discovered to be absent when subject to proper controls a priori.

It is quite disturbing to hear that this concern has not been raised previously.

Linda

 

[Ivor the Engineer]

Passing this along for Damien:

===========

Ivor writes: "Perhaps it's just my closed-minded engineering brain, but all I'm seeing in these results is evidence that some of the participants sucked at drawing numbered ping-pong balls at random out of a bag."

This is so remarkably perverse that I must regard it as heavy-handed irony or sarcasm. On reflection, though, perhaps it can be read as a re-statement of a truism: Information = -(Entropy)

Or did Ivor really *mean* it? Rather like: "All I'm seeing in the Olympic 100 meter results is evidence that some runners suck at falling down and breaking their legs or haring off in the wrong direction with their pants pulled over their heads."

Ivor: "Thus what is required is to alter various aspects of the experiment (e.g., bags, balls and other factors) until the draw frequencies match what the statistical model predicts they should be."

Great plan! You could also test the hypothesis that sighted people can see well enough to walk an obstacle course by shutting all the lights off, until the number of accidents and concussions matches what the statistical model of total blindness predicts they should be.

My point was we already know how to build very good random number generators, and all these experiments are evidence of is the set up being used is *not* a very good random number generator, most likely because of one or more of a number of mundane factors.

What I would suggest is modifying the experiment until the bias disappears. For example, change the bags to plastic jars with screw lids, the ping-pong balls to marbles and have the participants shake the containers in a predefined way for a minimum period of time to mix up the marbles. Each trial would be recorded on video to allow confirmation that the procedure was adhered to and to allow the results to be independently checked.

 

[Wowbagger]

It's too late to edit my last post, but here are more items you could add to the list of things that can't be ruled out, in Ertel's experiment:

(continued from the list in post #53)
* Statistical clustering effect (this is important, because if the results get "more even" over time, instead of more striking, this could be what you were really seeing.)
* A "Sokal"-like hoax (on part of those writing up the experiment)
* A Well-Organized Conspiracy to trick the experimenters

Not that some of those are likely. But, from the perspective of mainstream science, you have proven nothing, since they have not been ruled out.

 

[Moochie]

Perhaps it's just my closed-minded engineering brain, but all I'm seeing in these results is evidence that some of the participants sucked at drawing numbered ping-pong balls at random out of a bag.

Thus what is required is to alter various aspects of the experiment (e.g., bags, balls and other factors) until the draw frequencies match what the statistical model predicts they should be. I.e. find out what makes the anomaly disappear. Only then can you sensibly form a hypothesis as to what was causing the anomaly in the first place.

This makes sense to me. The only thing anomalous to me is how anyone can call this "science."


M.

 

[Limbo]

Another response to pass along. Prof Ertel writes:

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

Wowbagger: Limbo, I would like you to ask Suitbert Ertel a very important
question, if you could:
Do you consider your work, that I commented on, to be of genuine scientific
value?
If he answers NO: I will then promise to leave Ertel alone, from now on; as
I would have no more interest in him, myself.

Ertel’s reply: I don’t expect anyone having interest in me, but I would be
happy seeing others having interest in the phenomena for which I have found
evidence that surprised me. As long as some unusual and currently
unexplainable phenomenon is being reported as existent by only one or a few
scientists, as long as it has not been acknowledged by the majority of
scientists, the work of that minority cannot be regarded as of scientific
value. The attribute “scientific” implies an acknowledgment of the reported
phenomena by the majority of representatives of science, at least of
pertinent specialized disciplines . But the history of science is replete
with examples showing that observations eventuallyof great scientific value
are often preceded by long periods of disregard or even rejection by members
of the science institution whose competence is being challenged without
proper reactions.

Wowbagger: If he answers YES: Then, I would then like you to ask Suitbert a
follow up question: What part of the scientific method do you not
understand?! … If someone does not have such a "new idea", then how is that
attitude supposed to be more "fertile"?

Ertel’s reply: The somewhat hidden meaning of these questions seems to
imply that parapsychologists are not interested in explanations. This is not
so. Archeologists are still faced with eight scripts which cannot yet be
deciphered. But they are extremely interested in grasping the written
information.

Wowbagger: I meant: A testable hypothesis that can boost or clarify the
"anomalous ability".
… How are we supposed to rule out the following?:

* Bad experimental design (unintended bias on certain balls, for example)
* Deliberate cheating
* Paredolia or Confirmation bias
* Random chance (Hey, it could happen...)
* Erroneous recording
* An idiometer type of effect
* Etc.

Ertel's reply: Wowbagger has apparently not read my published reports about
my “grab bag” test * as he called it * a nice short term. He is right that
most tentative non-psi explanations can and should be tested
experimentally. I think I have been testing the relevant hypotheses that he
has listed here. The Just-Drawn-Ball bias figured by fls should be added.
None of these tests showed alternative effect indications. My references:

Ertel, S. (2005). "Psi test feats achieved alone at home: Do they disappear
under lab control?" Australian Journal of Parapsychology 10(2): 149-164.

Ertel, S. (2005). "Are ESP test results stochastic artifacts?" Journal of
Consciousness Studies 12(3): 61-80.

Ertel, S. (2005). The ball drawing test: Psi from untrodden ground.
Parapsychology in the twentieth century. M. A. Thalbourne and L. Storm.
Jefferson NC: McFarland: 90-123.

Ertel, S. (2007). "Außersinnliche Wahrnehmung unter der Kontrolle
organisierter Skeptiker." Zeitschrift für Anomalistik 7(3): 236-269.

P. S. Thanks to those who express appreciation of our effort at clarifying
the controversial issues.