Pigasus Awards & Sheldrake

[Ersby]

Excuse the cut and paste:

I looked at Sheldrake's experiment with Jay-Tee a while ago and I looked at Sheldrake's data, but this time I put it into chronoligical order, and I noticed something quite peculiar about the first seven sessions:

19 nov 96 - 18
11 dec 96 - 17
11 feb 97 - 16
19 mar 97 - 13
25 mar 97 - 13
7 may 97 - 12
1 jul 97 - 12
9 jul 97 - 17
29 aug 97 - 12
10 sep 97 - 12
21 sep 97 - 20
8 oct 97 - 15

(The number is the period when the owner began to return home)

Despite being chosen randomly, they decrease gradually from the first very long session to the seventh short one. Since the trials are spread out over eight months, there's no way the dog could recognise this, but the parents who were looking after her may have been thinking "she should be back by now", and subtely acting accordingly, which then effected the dog's behaviour.

It's certainly curious that the first seven trials should follow such a distinct pattern, and it should be amongst these that the best results are. The control sessions - where the owner did not return at all, unknown to the parents - did not happen until after the seventh trial.

Also, it would have been nice to see the data from the 50 sessions with Jay-Tee alone, since that would cancel out any chance of the dog reacting to body language. I did email Rupert Sheldrake asking to see it but despite an email from Pam Smart saying she'd pass on the request, I never heard anything.

 

[calebprime]

I want to be sure I understand.

The dog is in the presence of people who know when the person is coming home?

If so, that makes the experiment invalid.

Is this the case?

 

[Beth]

Excuse the cut and paste:

I looked at Sheldrake's experiment with Jay-Tee a while ago and I looked at Sheldrake's data, but this time I put it into chronoligical order, and I noticed something quite peculiar about the first seven sessions:

19 nov 96 - 18
11 dec 96 - 17
11 feb 97 - 16
19 mar 97 - 13
25 mar 97 - 13
7 may 97 - 12
1 jul 97 - 12
9 jul 97 - 17
29 aug 97 - 12
10 sep 97 - 12
21 sep 97 - 20
8 oct 97 - 15

(The number is the period when the owner began to return home)

Despite being chosen randomly, they decrease gradually from the first very long session to the seventh short one. Since the trials are spread out over eight months, there's no way the dog could recognise this, but the parents who were looking after her may have been thinking "she should be back by now", and subtely acting accordingly, which then effected the dog's behaviour.

It's certainly curious that the first seven trials should follow such a distinct pattern, and it should be amongst these that the best results are. The control sessions - where the owner did not return at all, unknown to the parents - did not happen until after the seventh trial.

Also, it would have been nice to see the data from the 50 sessions with Jay-Tee alone, since that would cancel out any chance of the dog reacting to body language. I did email Rupert Sheldrake asking to see it but despite an email from Pam Smart saying she'd pass on the request, I never heard anything.

Please post if you get anything from them. I would be interested to see your analysis. You do good work.

 

[tkingdoll]

Not really. Have you actually read his response to their criticisms?

Yes I have. And the response to that.

But I believe Wiseman's interpretation is the accepted one. Science has not rolled over and accepted the existence (or the possibility of the existence) of a telepathic dog.

Hence Sheldrake a) has no Nobel Prize and b) there isn't panic in the labs at the discovery that so much of what we thought we knew is wrong. It all just ticks along nicely, having a giggle at Sheldrake's para stuff on the way.

 

[Beth]

Yes I have. And the response to that
But I believe Wiseman's interpretation is the accepted one. Science has not rolled over and accepted the existence (or the possibility of the existence) of a telepathic dog.

Hence Sheldrake a) has no Nobel Prize and b) there isn't panic in the labs at the discovery that so much of what we thought we knew is wrong. It all just ticks along nicely, having a giggle at Sheldrake's para stuff on the way.

I certainly wouldn't expect science to roll over, panic and/or award a nobel prize on the basis of one unreplicated experiment. But a lack of acceptance is not the same as either a) debunking the work or b) establishing dishonesty on the part of the researcher. I would be interested in seeing Wiseman's response to Sheldrake's rebuttal if you could provide a link to it.

 

[Invidious]

Found this (response to response) in an archival thread:
http://www.luckfactor.co.uk/papers/psychicdogreply.pdf

Here's the old thread:
http://www.internationalskeptics.com/forums/archive/index.php/t-38499.html

 

[Beth]

Found this (response to response) in an archival thread:
http://www.luckfactor.co.uk/papers/psychicdogreply.pdf

Thank you. While I wouldn't say it was a complete rebuttal, it certainly illuminates many of the problems inherent in such research.

 

[bfinn]

That's exactly why the phenomenon doesn't exist.

Why would people be more receptive to those they know, and not guess e.g. that it was some telemarketer, wrong number, or whoever unknown might be calling?

I don't know (though I could speculate - e.g. it's not that surprising that we might react differently and with more certainty to people we know & recognise than to people we don't know). To find out you'd need to actually do the research.

There seems to be a strange mindset at work on this forum which says 'I can't think how something could be the case, so it can't be the case, so we shouldn't test it'. (I believe 'argument from lack of imagination' is the technical term.) As I've indicated before, this mindset seems more akin to religious faith than to scientific method.

That doesn't happen. No, people report that they think of someone they know, and whoopsie, they call.

Which is what we call "confirmation bias".

I take it you're referring to the anecdotal evidence that motivated the experiment, rather than the experiment itself. No-one denies that confirmation bias may at least partly explain the anecdotal evidence. The experiment is naturally intended to remove confirmation bias.

 

[Invidious]

There seems to be a strange mindset at work on this forum which says 'I can't think how something could be the case, so it can't be the case, so we shouldn't test it'. (I believe 'argument from lack of imagination' is the technical term.) As I've indicated before, this mindset seems more akin to religious faith than to scientific method.

Wow. A "strange mindset at work on this forum." Diabolical. Is it that crazy, unimaginative mindset where they demand evidence for every damned little thing? I have it on good authority (God's) that that's what they refer to as skepticism.

Anyway, big forum, sweeping generalization, I have 2 sugar lumps which will clear right up that case of hyperbole you've got. Take them every four hours with a cold shower, and post again in the morning.

 

[Ersby]

I want to be sure I understand.

The dog is in the presence of people who know when the person is coming home?

If so, that makes the experiment invalid.

Is this the case?

The dog is in the presence of people who could deduce when the owner is coming home, at least during those first seven trials. The first trial was very long - in fact almost the longest it could be given the protocol. So the parents would've been expecting the owner to come back at least by that time. And so as the time slowly decreased, so the parents may have started thinking "she should be back round about now", and the dog may have picked up on their subconscious cues.

Richard Wiseman recently did an interview on a podcast, Skeptiko, in which he mentioned the Jay-tee debate. He said, iirc, that the experiments set out to measure two different things. Sheldrake was interested if the dog seemed to know when the owner arrived home, and Wiseman wether the dog seemed to know when the owner decided to come home. Wiseman also said that he agreed that the data he got does seem to accord with Sheldrake's.

http://www.skeptiko.com/index.php?id=18

I'd need to look a bit closer, but from reading the pdf by Wiseman posted in this thread, and looking at the graphs in Sheldrake's paper, they don't seem to match up re. amount of time the dog stayed at the window. But that's just a quick glance. I'd need to sit down and work it out properly, and I'm not sure if I have enough time right now.

 

[Ersby]

Please post if you get anything from them. I would be interested to see your analysis. You do good work.

Thanks. I emailed this request some months ago so I don't expect a reply on this subject any time soon. It may have got lost - he answered my other email (on another subject) quite promptly.

 

[CFLarsen]

I don't know (though I could speculate - e.g. it's not that surprising that we might react differently and with more certainty to people we know & recognise than to people we don't know). To find out you'd need to actually do the research.

See later.

There seems to be a strange mindset at work on this forum which says 'I can't think how something could be the case, so it can't be the case, so we shouldn't test it'. (I believe 'argument from lack of imagination' is the technical term.) As I've indicated before, this mindset seems more akin to religious faith than to scientific method.

Nonsense. Skeptics have plenty of imagination, but what they won't do, is merely imagine without trying to find out if whatever is imagined is actually there.

Your reaction is common to all Superstitious: You are simply not used to the kind of scrutiny that you get from skeptics, so you think it must be the fault of the skeptics.

I take it you're referring to the anecdotal evidence that motivated the experiment, rather than the experiment itself. No-one denies that confirmation bias may at least partly explain the anecdotal evidence. The experiment is naturally intended to remove confirmation bias.

No, that's exactly what it hasn't. If they wanted to remove confirmation bias, they would have included unknown people as callers, too. If the receiver could determine if it was a known person or an unknown person, we would immediately know if the phenomenon distinguished between the two kinds.

Yet another simple test that wasn't done.

 

[Zep]

This doesn't make a protocol using only 4 callers known to the recipient an invalid test of telepathy.

Actually, that is precisely a situation that does invalidate such a protocol. To test using people known to the recipient is to invite collusion with an open door, deliberate or otherwise.

If someone claims they know telepatically when a friend of family member is about to call them then test them using 4 known people, or test them using 100 unknown and 4 known. It won't make a difference to the validity of the test.

see above

Since Sheldrake states that this apparent phenomenon usually occurs when the caller is known to the recipient it would make good sense to test using known callers. The problems you mention could be controlled for. Testing with known callers really isn't a problem, which is why Chris French is attempting a collaborative replication with tighter protocols also using known callers.

I imagine you would need to compare that with the same testing procedure run with callers unknown to the recipient...to see if there really IS something, or if it is just some form of collusion. Which is what I suggested initially.

 

[davidsmith73]

Actually, that is precisely a situation that does invalidate such a protocol. To test using people known to the recipient is to invite collusion with an open door, deliberate or otherwise.

You can prevent collusion with the correct experimental procedure.

I imagine you would need to compare that with the same testing procedure run with callers unknown to the recipient...to see if there really IS something, or if it is just some form of collusion. Which is what I suggested initially.

If you've controlled for collusion, then you can't turn to it as an explanation if trials with known callers produce results and unknown callers do not. If you did, then you are admitting that you actually didn't control for collusion at all! That's not how science works.

 

[bfinn]

Nonsense. Skeptics have plenty of imagination, but what they won't do, is merely imagine without trying to find out if whatever is imagined is actually there.

I am all in favour of trying to find out, i.e. conducting experiments. My comment was a response to the following:

That's exactly why the phenomenon doesn't exist.

Why would people be more receptive to those they know, and not guess e.g. that it was some telemarketer, wrong number, or whoever unknown might be calling?

I.e. 'I can't think of a way it could work, so it doesn't work'. Which is not arguing from experimental results (a posteriori), it is the exact opposite (a priori).

And see tking's various earlier posts saying repeatedly 'We shouldn't be testing this because we know in advance it won't work'. Same mindset.

(Philosophers present will know that the problem of induction may entail that we can't know the outcome in advance of any experiment, no matter how mundane.)

 

[bfinn]

If they wanted to remove confirmation bias, they would have included unknown people as callers, too. If the receiver could determine if it was a known person or an unknown person, we would immediately know if the phenomenon distinguished between the two kinds.

The test was not primarily trying to tell whether the recipient can distinguish between known and unknown people. It was trying to tell (more specifically) whether the recipient can identify which person is calling.

Are you suggesting that the latter would not count as telepathy, and only the former would?!

 

[davidsmith73]

No, that's exactly what it hasn't. If they wanted to remove confirmation bias, they would have included unknown people as callers, too. If the receiver could determine if it was a known person or an unknown person, we would immediately know if the phenomenon distinguished between the two kinds.

Yet another simple test that wasn't done.

(actually, unknown callers were included in the experiment)

Including unknown, known, animals or the speaking clock. It doesn't matter as long as the correct controls are in place.

 

[kieran]

You can prevent collusion with the correct experimental procedure.



If you've controlled for collusion, then you can't turn to it as an explanation if trials with known callers produce results and unknown callers do not. If you did, then you are admitting that you actually didn't control for collusion at all! That's not how science works.

I'd agree with this only with the following one-word modification: If you've successfully controlled for collusion, then you can't turn to it as an explanation if trials with known callers produce results and unknown callers do not.

I don't want to be pedantic here but claiming you have controlled for collusion, and actually doing it aren't the same thing. Science works by re-examing itself repeatedly, and refining/correcting when appropriate and necessary. Would it be wrong to check the effectiveness of your controls?

 

[davidsmith73]

I'd agree with this only with the following one-word modification: If you've successfully controlled for collusion, then you can't turn to it as an explanation if trials with known callers produce results and unknown callers do not.

I don't want to be pedantic here but claiming you have controlled for collusion, and actually doing it aren't the same thing. Science works by re-examing itself repeatedly, and refining/correcting when appropriate and necessary. Would it be wrong to check the effectiveness of your controls?

Of course not, to your last question. My argument was assuming that successful controls had been established.

 

[sinnikal]

Aah! Assumption. That great leveller