Dean Radin - harmless pseudo-psientist.

humber said:
Yes, I agree. But the error may be more subtle than that. If there is such an error, then the HD may transmit that information.
Click to expand...

Is there any information on the spec of the PC being used and, more importantly, is the total number of photos in the pool specified anywhere (I didn't see it, just wondering)?

The reason I ask is that although the HD noise represents theoretical leakage it would be totally inconsequential in all but the very oldest of equipment, because modern computers have enough resource not to have to access the HD every time and even if they do, they would give away no information because the access is so quick and quiet.

The number of images in the pool is important because if there were a large number then the HD noise would be irrelevant (even if it was consequential) because a single subject would be unlikely to encounter the same image twice.

Indeed, it could be argued that experimental design should (perhaps must) include functionality to prevent any one image being shown twice, because surely that would affect the response, whether pro-active or re-active. (The second time I see an icky image I'm unlikely to respond as strongly as the first time).

I still agree that the HD issue should be eliminated, if only to avoid these discussions. However, at this moment I see it as no more a factor than the subjects learning to predict images by detecting the temperature difference in the room due to CPU temperature variance and case fan speed variation.
 
humber said:
Yes, I agree. But the error may be more subtle than that. If there is such an error, then the HD may transmit that information. Perhaps I am reading too much into this, but the randomisation is only tested at the RNG level, and not through to the actual picture displayed? The entiresystem as it were?
Click to expand...

The paper says that "examination of the actual sequence of targets used in these experiments showed that the autocorrelations to lag 15 were all in alignment with chance expectation" (p268), although the data is not shown. I'm not sure what the autocorrelation function does though.

Perhaps the HD issue is unimportant, but it does go to the author's thoroughness, and one of many other examples where leakage has been given lip-service, but not eliminated.
Why not make the HD inaudible, and remove all doubt ?
Click to expand...

I think the point is that the disk noise is inaudible during the presentiment period because the HD has not been accessed at that point. Any argument about inferring the sequence of pictures based on HD noise from trial to trial is equally applicable to information gained from simply remembering the sequence of emotional/calm pictures. Therefore, I would say there's no point in making the HD inaudible because the experiment can hardly eliminate memory for the sequence of pictures. It really just comes down to adequate randomisation.

Unless, all of the presentiment effect happens during the HD access period. But this is highly unlikely since the presentiment effect can be seen to occur well before this period.
 
I've got a 2 year-old Dell Pentium 4 at work with an 80GB Hard drive and I can hear the heads move, even with the background noise in the office. The first trial was using a 66MHz Pentium, so it probably had a hard drive smaller than 10GB. The experiment was performed in a quiet room.

What if the calm and emotional pictures were roughly split between two separate areas on the disk? The sound of head movement would indicate a likely change from a calm to emotional picture or vice versa?
 
baron said:
Is there any information on the spec of the PC being used and, more importantly, is the total number of photos in the pool specified anywhere (I didn't see it, just wondering)?

The reason I ask is that although the HD noise represents theoretical leakage it would be totally inconsequential in all but the very oldest of equipment, because modern computers have enough resource not to have to access the HD every time and even if they do, they would give away no information because the access is so quick and quiet.
Click to expand...


Surely HD noise is not a leakage issue if the presentiment analysis is done on SCR data before the HD is accessed in each trial?

The number of images in the pool is important because if there were a large number then the HD noise would be irrelevant (even if it was consequential) because a single subject would be unlikely to encounter the same image twice.
Click to expand...

120 images in most of Radin's experiments I think - 80 calm, 40 emotional

Indeed, it could be argued that experimental design should (perhaps must) include functionality to prevent any one image being shown twice, because surely that would affect the response, whether pro-active or re-active. (The second time I see an icky image I'm unlikely to respond as strongly as the first time).
Click to expand...

In all the experiments, if an image was selected that had been selected before, it was rejected and a new image was randomly chosed. However, the randomisation was with replacement.
 
davidsmith73 said:
I think the point is that the disk noise is inaudible during the presentiment period because the HD has not been accessed at that point. Any argument about inferring the sequence of pictures based on HD noise from trial to trial is equally applicable to information gained from simply remembering the sequence of emotional/calm pictures. Therefore, I would say there's no point in making the HD inaudible because the experiment can hardly eliminate memory for the sequence of pictures. It really just comes down to adequate randomisation.

Unless, all of the presentiment effect happens during the HD access period. But this is highly unlikely since the presentiment effect can be seen to occur well before this period.
Click to expand...

No, no, no.

David, you have no idea what you are talking about.

First of all, two different images of the same dimensions - say, 1024x768 - can be of vastly different size, and therefore take very noticeable different time to retrieve. A very complex JPG of a train can be many, many times bigger than a simple JPG of a blue sky with few clouds.

Second, it is by far not certain that the computer accesses the hard disk for the pictures every time - it can also retrieve them from cache, either on the hard disk or from RAM.

Third, the program can access the pictures in many ways: It can force-read bit-by-bit on the hard disk, or it can read ahead before the experiment starts and have them ready in the memory (RAM). Even so, it still doesn't necessarily ensure that the pictures stay in RAM.

Whenever a computer is involved in an experimental setup, there is every reason to look deep into the possibilities of influencing the experiment.
 
davidsmith73 said:
<snip>

Unless, all of the presentiment effect happens during the HD access period. But this is highly unlikely since the presentiment effect can be seen to occur well before this period.
Click to expand...

You are assuming the same cause for the effect seen in each different experiment. My comments are with respect to experiment 1, where the hard disk access was at the start of the pre-stimulus period.

What I find annoying about Radin's work is so much changes between experiments. When you point out one flaw in an experiment it is dismissed because a different experiment, which has other possible flaws, shows an effect.

He is attempting to show a tiny effect, yet keeps on changing the protocol of the experiment. Each change could effect the result.
 
Last edited:
Ivor the Engineer said:
I've got a 2 year-old Dell Pentium 4 at work with an 80GB Hard drive and I can hear the heads move, even with the background noise in the office. The first trial was using a 66MHz Pentium, so it probably had a hard drive smaller than 10GB. The experiment was performed in a quiet room.
Click to expand...

For a 66MHz PC then yes, the HD would make a noise like a bag of bolts and there wouldn't be enough memory to store the photos in memory as they are retrieved, thereby necessitating the script to go back to disk on every retrieval.

So yes, in that instance, although I don't believe it's possible to interpret those sounds, I think they should have been eliminated for completeness.

Ivor the Engineer said:
What if the calm and emotional pictures were roughly split between two separate areas on the disk? The sound of head movement would indicate a likely change from a calm to emotional picture or vice versa?
Click to expand...

For a similarly slow PC, and in theory, if Calm images are located on one area of the disk and Activating images on another then it's possible there would be an difference in disk access times, and therefore noise. However, in practice files are very rarely, if ever, stored as a contiguous set of bytes, they are spread out over the disk as a result of file fragmentation. A single image may have hundreds or even thousands of fragments dotted all over the HD. The location of these fragments is not predictable.
 
Ivor the Engineer said:
You are assuming the same cause for the effect seen in each different experiment. My comments are with respect to experiment 1, where the hard disk access was at the start of the pre-stimulus period.

What I find annoying about Radin's work is so much changes between experiments. When you point out one flaw in an experiment it is dismissed because a different experiment, which has other possible flaws, shows an effect.

He is attempting to show a tiny effect, yet keeps on changing the protocol of the experiment. Each change could effect the result.
Click to expand...

That's a hallmark of pseudoscience, especially in parapsychology.

They keep talking about "replication", but they don't replicate. They tweak and twist.
 
davidsmith73 said:
I see now, sorry I didn't read your post properly. Let me just check if I've got this right.

The SCR will be huge after a subject views an emotional picture. This means that at the end of an emotional trial (I'm talking about Radin's design here) the rate of change of SCL may be greater than at the end of a calm trial and may continue into the start of the next trial. Assuming this happens, then when the first SCL of the next trial is clamped to zero, there will be a greater change in the SCL during the pre-stimulus period for trials succeeding an emotional trial compared to a calm trial. Is that right so far?
Click to expand...

Yes.

So, even if there are twice as many calm photos then I don't see how this would produce an artifact. The probability of a calm trial after viewing an emotional picture is the same as the probability of a calm trial after viewing a calm picture so the effect should cancel out shouldn't it?
Click to expand...

Rather than effects that cancel each other out (then tend to all go in the same direction, after all), I'm thinking that most of the time there is little effect and occasionally there is a large effect. The residual effect of seeing an emotional photo is twice as likely to be subsequently associated with a calm photo as with an emotional photo (although the way the analysis is set up, the subject may be equally likely to be placed in the 'calm' or 'emotional' group, since the 'emotional' group contains subjects exposed to calm photos). If the occasional presence of a large effect is distributed unevenly between the two groups, then it will lead to a tiny, inconsistent, but sometimes significant, difference in the averages between the two groups (which is what we see with the different studies). I'd be interested in seeing what happens if you take this into account.

Linda
 
Ivor the Engineer said:
You are assuming the same cause for the effect seen in each different experiment. My comments are with respect to experiment 1, where the hard disk access was at the start of the pre-stimulus period.
Click to expand...

My comments were in reply to humber with regard to the experiments that controlled for HD noise by accessing the HD after the presentiment analysis period, i.e., experiments 2-4. However, I was just wondering if the SCR analysis period overlapped very slightly with the period when the HD was accessed. In other words, was the SCR sampled right up to the moment the image was displayed, or did sampling stop before the HD was accessed? It's not clear from the paper. Of course, you are right that the HD noise argument applies to an experiment where the disk is accessed at the start of the presentiment period. However, I'm not so sure the HD was accessed at the start of the presentiment period in experiment 1 (starting bottom of page 267):

"If the computer’s hard disk retrieved the target photo immediately after the button was pressed to begin a trial, and if the calm targets differed from the emotional targets either in terms of where they were located on the disk or their size, then the participant might have learned to associate the computer’s hard disk sounds with different upcoming targets. To avoid such possibilities, in all experiments the targets were not retrieved off the hard disk until immediately before they were displayed. Also, recall that in Experiments 2 through 4 the targets were not even selected until immediately before they were displayed."

So experiment 1 selected a target at the start of the trial but the image file was not accessed until immediately before presentation of the picture.

Does that mean no information about upcoming pictures could be inferred from HD noise in any of these experiments?

If so we can stop talking about it as a flaw in the experiment.
 
davidsmith73 said:
Surely HD noise is not a leakage issue if the presentiment analysis is done on SCR data before the HD is accessed in each trial?
Click to expand...

I was going to say "obviously", but on re-reading my own posts it looks like I've been talking rubbish on this issue. That's what comes of concentrating on trivia.

Right. HD noise cannot be an issue. That's my stance :)

davidsmith73 said:
120 images in most of Radin's experiments I think - 80 calm, 40 emotional

In all the experiments, if an image was selected that had been selected before, it was rejected and a new image was randomly chosed. However, the randomisation was with replacement.
Click to expand...

Right, thanks.
 
CFLarsen said:
First of all, two different images of the same dimensions - say, 1024x768 - can be of vastly different size, and therefore take very noticeable different time to retrieve. A very complex JPG of a train can be many, many times bigger than a simple JPG of a blue sky with few clouds.

Second, it is by far not certain that the computer accesses the hard disk for the pictures every time - it can also retrieve them from cache, either on the hard disk or from RAM.

Third, the program can access the pictures in many ways: It can force-read bit-by-bit on the hard disk, or it can read ahead before the experiment starts and have them ready in the memory (RAM). Even so, it still doesn't necessarily ensure that the pictures stay in RAM.

Whenever a computer is involved in an experimental setup, there is every reason to look deep into the possibilities of influencing the experiment.
Click to expand...


But if the pictures are retrieved after the presentiment analysis period, then the precise way they are retrieved does not matter, yes?

Unless you are saying that there is still the possibility that the pictures are accessed during the presentiment period?
 
davidsmith73 said:
But if the pictures are retrieved after the presentiment analysis period, then the precise way they are retrieved does not matter, yes?

Unless you are saying that there is still the possibility that the pictures are accessed during the presentiment period?
Click to expand...

I am saying that we don't know when the pictures are retrieved from the hard disk. That means we can't rule that factor out. Therefore, it's a possible leak. Agree?

Can you explain what a "calm" picture is? What is an "emotional" picture?
 
davidsmith73 said:
<snip>

So experiment 1 selected a target at the start of the trial but the image file was not accessed until immediately before presentation of the picture.

Does that mean no information about upcoming pictures could be inferred from HD noise in any of these experiments?

If so we can stop talking about it as a flaw in the experiment.
Click to expand...

Yes, but experiment 1 was still a test of clairvoyance, not presentiment.:p

So that leaves randomisation, 2:1 ratio of calm to emotional trials and physiological effects.

fls said:
Rather than effects that cancel each other out (then tend to all go in the same direction, after all), I'm thinking that most of the time there is little effect and occasionally there is a large effect. The residual effect of seeing an emotional photo is twice as likely to be subsequently associated with a calm photo as with an emotional photo (although the way the analysis is set up, the subject may be equally likely to be placed in the 'calm' or 'emotional' group, since the 'emotional' group contains subjects exposed to calm photos). If the occasional presence of a large effect is distributed unevenly between the two groups, then it will lead to a tiny, inconsistent, but sometimes significant, difference in the averages between the two groups (which is what we see with the different studies). I'd be interested in seeing what happens if you take this into account.
Click to expand...

Linda, can you explain the bolded section. I'm not sure I understand what you mean by "may equally likely to be placed in the 'calm' or 'emotional' group"?
 
You must log in or register to reply here.

Back
Top Bottom