Prestimulus Response Experiments?

[roger]

Here is a paper by Broughton on a failure to replicate:

http://72.14.207.104/search?q=cache...f+radin+presentiment&hl=en&gl=us&ct=clnk&cd=2(snip url)

~~ Paul

An interesting paper, perhaps. I haven't read it yet, but see in the abstract that they tackle the retest question. That's another issue with the papers above. They talk about strong performers, but never prove it. A person is a strong performer if they can show a positive correlation across multiple experiments. These papers are all taking the strong performers from a single experiment. The law of large numbers tells us there will always be strong performers within a single experiment. Any analysis that follows from those selected groups is immediately worthless. Another objection to the 13 person subset used to test the anticapatory strategy.

 

[davidsmith73]

He asserts that, but does not prove that, to my understanding. Could you explain why this would be true?

There is a specific hypothesis under test here - that an anticipatory strategy (just like the one you described previously) is the mechanism by which a significant difference in SCR between calm and emotional trials is generated. If his set of data is selected on the basis of a significant difference between calm and emotional trials then this dataset is the most likely one to display the anticipatory effect if it is present because such a strategy predicts such a significant difference, according to the hypothesis under test . I think its fair to say that no strategy was found to be present in his subset of data.

I offered a counterexample that suggests that it would be less likely to show this effect.

I address this below

Untrue. Uniform random data will show exactly that in some subsets. Other subsets will show nothing, and yet other subsets will show a negative correlation.

Yes, you are right. I'll address this below again.

Assume 4 trials, with random data. 4 so I can write out all the possibilities - obviously a real experiment will require more trials. I simplify by rating each trial as calm or excited, rather than the continuum that is actually measured.

With random data we generate (c = calm, e = excited):

CCCC
CCCE
CCEC
CCEE
...
EEEE

When matched against a specific picture series, some of those show no correlation, some positive correlation, and some negative. I'm sure we agree on that point. But, as far as the non-psi anticipatory strategy goes, some of those will show the anticipatory effect, and some won't.

Agreed so far. ETA: apart from the last sentence!

However, the better the match, the less of that effect it will show, I argue.
For example, suppose the picture series was CCCE. Take the data set CCCE. That will test low for the antipatory practices, because there is no build up from calm to excited. yet it is the highest possible performace possible for that picture series.

Your example can't show any form of anticipatory strategy because you are forcing the SCR dataset into only two discrete responses - C or E. The anticipatory strategy hypothesis predicts a gradual increase in SCR along calm trials, culminating in a maximum SCR at the emotional trial, after which the level of SCR response is then "re-set". I'm still a little unsure of what you are trying to show in your counter example.

Radin, by selecting the 13 highest performers, is quite probably selecting against the anticipatory strategy, not for it.

I don't see how this is the case at all. Your random data example is not a valid model for what the anticipatory hypothesis predicts, ie, it is not possible for your example to show such a strategy at all because you only have two discrete reponses.

The point being that the anticipatory strategy will only yield a modest positive correlation; random data can (and will) yield higher correlations, just by the law of large numbers.

I don't understand. A modest correlation between what?

 

[roger]

I don't see how this is the case at all. Your random data example is not a valid model for what the anticipatory hypothesis predicts, ie, it is not possible for your example to show such a strategy at all because you only have two discrete reponses.

As I said, I simplified the notation just so I could easily write it down. Infer for each C and E in my notation a continuous, analog signal. After all, to judge the results the experimenter has to decide if the multisecond, analog measurement corresponds to a calm or exciting picture.

My point is that one can posit an analog data set where it shows high correlation with the pictures, but doesn't show the anticipatory response.
I.e. for the CCCE picture sequence, assume an analog signal with remains pretty much at baseline, with no significant rises, until 3 seconds before the fourth picture, where it shows a stong spike. And assume a second analog signal with slowly rises from baseline until it reaches it's maximum, at 3 seconds before the fourth picture. Both show positive correlation to the CCCE picture sequence, however the first data set is a stronger correlation because it does not have that rising slope. The first data set also does not show the anticipatory response, because it doesn't have a rising slope prior to the fourth picture.

And that's the point. Random data can and will produce a stronger correlation than can the well documented anticipatory response, and it will not show the anticipatory response.

By choosing the top 13 performers, we risk selecting random data that fits that criteria. That it doesn't show evidence for the anticipatory response is not proof that the data is due to PSI, since we would expect exactly that result if the data was random anyway.

I don't understand. A modest correlation between what?

Between the recorded data and the picture sequence.

 

[davidsmith73]

My point is that one can posit an analog data set where it shows high correlation with the pictures, but doesn't show the anticipatory response.
I.e. for the CCCE picture sequence, assume an analog signal with remains pretty much at baseline, with no significant rises, until 3 seconds before the fourth picture, where it shows a stong spike. And assume a second analog signal with slowly rises from baseline until it reaches it's maximum, at 3 seconds before the fourth picture. Both show positive correlation to the CCCE picture sequence, however the first data set is a stronger correlation because it does not have that rising slope. The first data set also does not show the anticipatory response, because it doesn't have a rising slope prior to the fourth picture.

I see what you mean now. That's a good point. But I'm still a little sceptical over this because he did analyse all his participants that achieved significance (according to his arbitrary emotional/calm picture rating of course). I would expect that some of these individuals would show an anticipatory strategy if such a strategy was responsible for the overall experimental results.

So would it be possible that the remaining 90 individuals in Radins showed an anticipatory strategy? It's not clear how Radin did his analysis. I think its implied that the remainder didn't achieve a significant difference between calm and emotional taken individually. However, would it be possible for an anticipatory effect to be present in the difference between calm and emotional trials when you average everyones response in some way. Seems quite over complicated.

It would be much better for this experiment if such a strategy could be ruled out in the experimental design. The only way I can think of is to keep the participants naive to the fact that an emotional trial will appear. Perhaps let them think they are involved in an experiment looking at normal responses to calm stimuli, give them a sequence of calm pictures, then place an emotional one in there without them knowing. Probably wouldn't be allowed for eithical reasons here in UK! We're a bit soft.

 

[roger]

It would be much better for this experiment if such a strategy could be ruled out in the experimental design. The only way I can think of is to keep the participants naive to the fact that an emotional trial will appear.

Well, that's where I was going with the long periods between individual trials, though you rightly pointed out a limitation with that strategy.

How about this? It'd take a long time to generate a meaningful data set, but I think it would work.

Each experiment shows 4 pictures. The first 3 are the same for all testees, and are rated as 'calm'. The last picture is randomly selected as either calm or stimulating. The only data analysed for each experiment is the 3-4 seconds prior to the fourth picture. The subgroup where the 4th picture is calm is your control group, and the excited picture is your experimental group. All people are blinded to the purpose of the experiment. Data analysis method is chosen prior to the data collection, and no post hoc reasoning, data analysis, etc., is allowed. (of course you can do that to decide what you want your next experiment is, but you can't draw conclusions from those post poc decisions). All collected data and analsysis methods are published and/or made available to researchers - nothing is hidden from "negative sceptics", or whatever term you want to use. The papers are peer reviewed in an open environment.

If they can repeatedly show statistical significance in that scenerio, I'd tenatively agree that there was something going on (tenative because I'm no experimental scientist, and I've probably missed something that needs to be tightened up).

 

[davidsmith73]

How about this? It'd take a long time to generate a meaningful data set, but I think it would work.

Each experiment shows 4 pictures. The first 3 are the same for all testees, and are rated as 'calm'. The last picture is randomly selected as either calm or stimulating. The only data analysed for each experiment is the 3-4 seconds prior to the fourth picture. The subgroup where the 4th picture is calm is your control group, and the excited picture is your experimental group. All people are blinded to the purpose of the experiment. Data analysis method is chosen prior to the data collection, and no post hoc reasoning, data analysis, etc., is allowed. (of course you can do that to decide what you want your next experiment is, but you can't draw conclusions from those post poc decisions). All collected data and analsysis methods are published and/or made available to researchers - nothing is hidden from "negative sceptics", or whatever term you want to use. The papers are peer reviewed in an open environment.

Good suggestion. Only Dick Biermann has done something similar to this. He was performing a presentiment experiment like the one from Radin's paper but he added an extra experiment before the main one. Participants were told they would be shown only a series of about 10 calm pictures in order to calclulate a baseline reading, but he slipped in an emotional picture too. All participants got at least 3 calm pictures, just like your suggestion, with the only essential difference that he sequentially randomised the placement of the emotional picture after the 3 calm ones. So the emotional one could be 4th or 6th or 9th etc, up to a limit of 10th place. I think he did this to further test the sequential anticipatory strategy even though the participants were naive (so we assume) to the rogue emotional picure. Obviously participants could only have one trial so he didn't have a large n number. He did not get a significant result but there was a visible rise in SCR before the emotional picure. Interesting and definately worth a replication attempt.

Found the experiment. See Study 2.

http://m0134.fmg.uva.nl/publications/1998/retrocausal_tucson3.pdf

If they can repeatedly show statistical significance in that scenerio, I'd tenatively agree that there was something going on

I agree

 

[Ersby]

just thinking out loud...

I'd say that of all parapsychology's claims, this one seems the most promising. The idea of people remote viewing and describing locations that are distant in both space and time? Forget it. The idea we subconsciously react to a stimuli in the immediate vicinity and in the immediate future is simply more plausible. To me, anyway.

Unfortunately, the results are not easily described to a layman, and tend to rely on statistics and technical details to a level beyond most people's understanding. Plus, I think it suffers from a certain "unsexiness" that things like remote viewing or the Global Consciousness Project don't have. They can talk about spies or our collective soul crying out at some disaster. The prestimulus response work doesn't have that option - it's pretty dry stuff, which is perhaps why there's not been a huge amount of work on the subject, despite its relative ease.