36 Academic Snobbery and the Journal of Scientific Exploration (Podcast)

[Chris Noble]

Originally Posted by Larry Boy

You're talking about the "thirty ordinary homecomings", which are analyzed independently of the twelve random homecomings. In the paragraphs above the one that you quoted, Sheldrake describes "a pre-planned series of 12 experiments with randomly selected return times", in which "Jaytee was left at PS parents' flat and PS did not know in advance when she would be returning." Why do you insist on lumping together random and non-random trials?

You were the one who mentioned these trials above.

[Wiseman, Smith & Milton] suggest that Jaytee may simply have gone to the window more and more the longer Pam was out, and hence been there most when she was on the way home. But a comparison of Jaytee's behaviour during Pam's short, medium and long absences shows that this was not the case...

The "random" trials weren't all that random.

Three of the 12 experiments were carried out in the afternoon, with beeps at 2:22, 3:04 and 3:36 PM; the remaining experiments were carried out in the evening with beeps at a range of times between 8:09 and 9:39 PM.

9 out of 12 had the beep between 8:09 and 9:39 PM. That's hardly random at all.

You will notice that in the preliminary investigations, the 30 "ordinary" homecomings and the 12 "random" homecomings there are two peaks in the early afternoon and in the late evening where the vast majority of homecomings take place.

In all of these trials there is a consistent pattern in the times at which the owner returned home. Labelling some of the trials "random" doesn't change the fact that they had the same pattern.

[Larry Boy]

Originally Posted by Chris Noble

You were the one who mentioned these trials above.

In my first post I mentioned the random trials, under "Results" and "Experiments with randomly-selected return times" in Sheldrake's paper "A Dog That Seems To Know When His Owner is Coming Home: Videotaped Experiments and Observations".
In my second post I mentioned, again, the random trials, plus "Pam's short, medium and long absences" and "control experiments in which Pam did not come home", which Sheldrake doesn't specify (he instead gives a reference to his book).
Then I just replied to your arguments.

So, how can you claim that I was the one that brought up the "thirty ordinary homecomings"?

Originally Posted by Chris Noble

The "random" trials weren't all that random.


9 out of 12 had the beep between 8:09 and 9:39 PM. That's hardly random at all.

It's random within that time frame. The experiments started around that time as well, it's not like all of them started in the morning and then ended at around the same time in the evening, in which case your criticism may have been valid.

"These 'beep' times were within a pre-arranged period, between 45 and 90 minutes long. This period commenced 80 minutes to 170 minutes after PS had gone out. The 'beep window' was then divided into 20 equal intervals, and one of these was selected at random by throwing a die three times, to determine the page, row and column in standard random number tables (Snedecor and Cochran, 1967). Reading downwards from this point looking at the first two digits of each random number, the first pair of digits between 01 and 20 determined the time at which the beep was to be given."

Originally Posted by Chris Noble

You will notice that in the preliminary investigations, the 30 "ordinary" homecomings and the 12 "random" homecomings there are two peaks in the early afternoon and in the late evening where the vast majority of homecomings take place.

That's because the vast majority of the experiments were conducted at those times!

Originally Posted by Chris Noble

In all of these trials there is a consistent pattern in the times at which the owner returned home. Labelling some of the trials "random" doesn't change the fact that they had the same pattern.

The pattern is due to the experiments being conducted at certain times during the day.

[Chris Noble]

Originally Posted by Larry Boy

In my first post I mentioned the random trials, under "Results" and "Experiments with randomly-selected return times" in Sheldrake's paper "A Dog That Seems To Know When His Owner is Coming Home: Videotaped Experiments and Observations".
In my second post I mentioned, again, the random trials, plus "Pam's short, medium and long absences" and "control experiments in which Pam did not come home", which Sheldrake doesn't specify (he instead gives a reference to his book).
Then I just replied to your arguments.

So, how can you claim that I was the one that brought up the "thirty ordinary homecomings"?

The "short, medium and long absences" are the "thirty ordinary absences".

It's random within that time frame. The experiments started around that time as well, it's not like all of them started in the morning and then ended at around the same time in the evening, in which case your criticism may have been valid.

"These 'beep' times were within a pre-arranged period, between 45 and 90 minutes long. This period commenced 80 minutes to 170 minutes after PS had gone out. The 'beep window' was then divided into 20 equal intervals, and one of these was selected at random by throwing a die three times, to determine the page, row and column in standard random number tables (Snedecor and Cochran, 1967). Reading downwards from this point looking at the first two digits of each random number, the first pair of digits between 01 and 20 determined the time at which the beep was to be given."



That's because the vast majority of the experiments were conducted at those times!



The pattern is due to the experiments being conducted at certain times during the day.

All of the statistical analyses assumed that the behaviour of the dog should be completely random. If there are patterns in the homecomings of the owner then it is hardly surprising that there are patterns in the behaviour of the dog.

The RPA calculations made the assumption that under the null hypothesis, Jaytee should have spent about the same amount of time at the window in each of the 10-minute periods

This assumption is unjustified. This assumes that the dog is completely oblivious to the patterns in the comings and goings of the owner, the behaviour of the owners parents and other information that was not controlled for. If the dog has even a rough idea of the typical time that its owner is expected to come home and starts going to the window around that time then the observed pattern will be reproduced. The dog is expected to have spent more time on average at the window in the period when its owner is coming home. The fact that in the set of 12 "random" trials the dog started the behaviour before the owner knew she was going to be returning suggests that the dog is doing this not with some unknown supernatural talents but with the totally ordinary processes of an intelligent animal.

The null hypothesis assumes that the dog is completely stupid. Why it is preferable for the dog to be stupid but telepathic is beyond me.

[Larry Boy]

Originally Posted by Chris Noble

The "short, medium and long absences" are the "thirty ordinary absences".

I mentioned it in passing, along with the random trials. What's your point? Does it justify your claiming that all the trials were non-random?

Originally Posted by Chris Noble

All of the statistical analyses assumed that the behaviour of the dog should be completely random. If there are patterns in the homecomings of the owner then it is hardly surprising that there are patterns in the behaviour of the dog.

There are no patterns in the homecomings, there are patterns in the times she leaves. How does the dog calculate when she will come home based on when she leaves, if the homecomings are randomly determined?

[Chris Noble]

Originally Posted by Larry Boy

I mentioned it in passing, along with the random trials. What's your point? Does it justify your claiming that all the trials were non-random?




There are no patterns in the homecomings, there are patterns in the times she leaves. How does the dog calculate when she will come home based on when she leaves, if the homecomings are randomly determined?

The "random" homecomings aren't that random.

They arranged around times where the owner is available to return home which coincide with peaks in the time when she routinely comes home.

The owner knows the period in which she will return home because she arranged these times. From the graphs the mean time for the beep into this "beep" period is ~27 minutes with a standard deviation of 11 minutes.

The dog only needs to have a rough estimate of the time when the owner is likely to return and to start going to the window a bit before this time. The dog will then sepnd more time at the window during the return period than the main period with no need for supernatural explanations.

[David Bailey]

Originally Posted by Chris Noble

The "random" homecomings aren't that random.

They arranged around times where the owner is available to return home which coincide with peaks in the time when she routinely comes home.

Why not paste up the specific data you are referring to, and let's all get to assess that claim.

David

[Larry Boy]

Originally Posted by Chris Noble

From the graphs the mean time for the beep into this "beep" period is ~27 minutes with a standard deviation of 11 minutes.

My understanding is that it was randomly determined also when the beep periods themselves were going to be. "These 'beep' times were within a pre-arranged period, between 45 and 90 minutes long. This period commenced 80 minutes to 170 minutes after PS had gone out."

You're making it sound as if the owner came home approximately the same time on every trial, but if you look at the graphs, the beep periods start at quite varying times, at one time 90 minutes into the trial, at another time 140 minutes into the trial, at yet another time 110 minutes into the trial, at another time 160 minutes into the trial, and so on. Almost every time, the peak of Jaytee's waiting behaviour corresponds to the homecoming of the owner, although this waiting behavior rarely lasts more than 5 to 10 minutes.

[Chris Noble]

Originally Posted by Larry Boy

My understanding is that it was randomly determined also when the beep periods themselves were going to be. "These 'beep' times were within a pre-arranged period, between 45 and 90 minutes long. This period commenced 80 minutes to 170 minutes after PS had gone out."

You're making it sound as if the owner came home approximately the same time on every trial, but if you look at the graphs, the beep periods start at quite varying times, at one time 90 minutes into the trial, at another time 140 minutes into the trial, at yet another time 110 minutes into the trial, at another time 160 minutes into the trial, and so on. Almost every time, the peak of Jaytee's waiting behaviour corresponds to the homecoming of the owner, although this waiting behavior rarely lasts more than 5 to 10 minutes.

My point is very simple. The periods where there was some randomisation were not very large. Only one was 90 minutes the rest were closer to 45 minutes. These periods had been prearranged to occur when the owner was available to come home. The majority occurred in a small time period which coincided with a peak in the normal routine homecomings as seen from the preliminary investigations.

The owner knew when the start of the beep period was going to occur. It is possible that this expectation was unconciously communicated to the dog. The only effect of the randomistaion was a standard deviation of about 11 minutes in the distribution of return times from the start of the beep period.

[Larry Boy]

Originally Posted by Chris Noble

My point is very simple. The periods where there was some randomisation were not very large. Only one was 90 minutes the rest were closer to 45 minutes.

You're forgetting that the beep periods commenced at some point during a time frame of 80 to 170 minutes after the start of the trial.

Originally Posted by Chris Noble

The majority occurred in a small time period which coincided with a peak in the normal routine homecomings as seen from the preliminary investigations.

"Three of the 12 experiments were carried out in the afternoon, with beeps at 2:22, 3:04 and 3:36 PM; the remaining experiments were carried out in the evening with beeps at a range of times between 8:09 and 9:39 PM."

Is one hour and a half a "small time period", when the homecomings didn't last more than about 15 to 20 minutes? How does the dog pick out the right time to sit by the window?

Originally Posted by Chris Noble

The owner knew when the start of the beep period was going to occur. It is possible that this expectation was unconciously communicated to the dog.

I'm open to the possibility, but how do you communicate unconsciously to a dog that tonight, you're going to start going home sometime during a time frame between 45 to 90 minutes commencing at so-and-so many minutes from when you're leaving?

Originally Posted by Chris Noble

The only effect of the randomistaion was a standard deviation of about 11 minutes in the distribution of return times from the start of the beep period.

Once again, you're forgetting that the beep periods commenced at some point during a time frame of 80 to 170 minutes after the start of the trial.

[Chris Noble]

Originally Posted by Larry Boy

You're forgetting that the beep periods commenced at some point during a time frame of 80 to 170 minutes after the start of the trial.

No, I'm not. If the owner is aware of ther range of time in which she is going to return then she can communicate this to the dog or if her parents are aware then they can communicate this to the dog. If the owner arranged the return period to coincide when she was going to be available to return home then this could also introduce some degree of her normal routine

"Three of the 12 experiments were carried out in the afternoon, with beeps at 2:22, 3:04 and 3:36 PM; the remaining experiments were carried out in the evening with beeps at a range of times between 8:09 and 9:39 PM."

Is one hour and a half a "small time period", when the homecomings didn't last more than about 15 to 20 minutes? How does the dog pick out the right time to sit by the window?

The dog generally starts going to the window before the owner starts to return. In many cases the dog started going 30 or so minutes before the routine time although in most cases it was a lot less.

The major point that I am making is that all the dog has to do for it to be spending more time at the window during the return time than during the main period is for it to start going to the window more and more often starting any time between the start of the recording up to when the owner starts the return journey. The null hypothesis that the dog will spend the same time at the window at all time periods assumes that it has no knowledge whatsoever about the likely return of the owner. I find it highly implausible that the dog is completely oblivious to the patterns in the comings and goings of its owner.

I'm open to the possibility, but how do you communicate unconsciously to a dog that tonight, you're going to start going home sometime during a time frame between 45 to 90 minutes commencing at so-and-so many minutes from when you're leaving?

How do owners unconciously communicate that they are about to take the dog for a walk? I don't think that you need telepathy or "morphic resonance" to explain this one either.

The dog obviously spend a large amount of time paying attention to the comings and goings of its owner. Whether the owner realises it or not she would have behaved differently if she was going to stay away for a night for example. The owners parents would have acted differently too. Do you think the dog could have picked this up? Sheldrake implicitly assumes that the dog cannot pick this up with its natural senses.

The way to tackle this phenomenon would be to design experiments that test dog's natural senses. How much can dogs pick up naturally?

Start with whether the dog can detect whether the owner is going to stay out for the night then work back to the amount of information the dog would need to reproduce the phenomena that Sheldrake recorded.