Thursday, September 25, 2008

Presentiment demos



I do an interview for a documentary film or TV show typically once or twice a month. Occasionally the director wants to film a demonstration of a test in the lab, and one of the options I offer is a presentiment experiment using a skin conductance level (SCL) measurement. About half the time the director or producer decides to do the test him or herself.

I noticed some years ago when I started doing these demo experiments that they often resulted in pretty good results, which I found surprising because the set and setting of interviews are not nearly as well controlled as actual experiments. But starting this year, out of curiosity, I decided to keep a closer eye on the outcomes of those demos.

The image here shows the composite results after running four presentiment demos this year, so far. This is not based on a selection of results -- this is all the data from the four demos. There was one demo per interview, and each demo consisted of 40 randomly selected images out of a pool of about 700 images, for a total of 160 targets of widely varying emotionality. The trials were partitioned based on a median split of the emotionality ratings of the 160 trials, so the 80 targets rated as being most arousing (based on the international standard rating per image) were defined as emotional, and the bottom 80 target were defined as calm.

Each trial was normalized to allow the emotional and calm trials to be easily pooled across all four sessions, so the y axis is in terms of z scores (standard normal deviates). The difference in the two curves prior to the stimulus (i.e., the moment the target is selected and shown, at second 0) is not quite statistically significant, mainly because with 160 trials there is insufficient statistical power. But the differential effect is in alignment with the results that I and others have obtained in previous experiments.

What it shows is that about 3 seconds before the stimulus is randomly selected, the SCL measurement begins to differentiate according to the future target selection. A few people have misinterpreted what's going on here. What presentiment predicts is that prior to a randomly selected emotional target SCL should become more aroused than prior to a randomly selected calm target. This only means that the ensemble emotional curve is predicted to be higher than the ensemble calm curve, and not that the emotional curve will necessarily go into positive territory.

39 comments:

Mind said...

Dr. Radin, the presentiment experiments are one the most promising experiments in the psi field, at least in the recent times. I've read about them in your Entangled Minds book and you often refer to it.

Yet, recently I've come across 2 different critical analysises of the experiment, at 2 places on the Internet. One was on JREF forum but look quite solid, using a simulated copmuter model at: http://forums.randi.org/showthread.php?t=123007

Another review of the original paper was done for our mind-energy.net forum at:
Review of: "Electrodermal Presentiments of Future Emotions".

I wonder what's you response to some of the critique and suggestion to alter the protocol, for example to use longer periods between pictures, since the relaxation trend is still continuing at this time.

Dean Radin said...

I've commented on both of those critiques on this blog. My colleagues and I have been aware of these and other proposed artifacts for many years, we've studied them, and they do not account for the observed results.

E.g., anticipatory models assume that the experiment uses dichotomous targets (only calm vs. emotional). But most of my presentiment experiments (including the one mentioned in this post) select targets at random from a large pool of pictures of widely varying emotionality. That avoids the gambler's fallacy type of strategy, which is what anticipatory models are based upon.

David Bailey said...

Dean,

Regarding anticipation strategy, didn't you also look for anticipation in the simpler dichotomous studies, and not find any - or not enough. This would seem quite decisive because you would obviously see anticipation building up over a series of calm images.

Can you say if the anticipation explanation produces an effect of the same size as you observe in your experiments?

The anticipation effect is very subtle and counter-intuitive - because at first sight no anticipation strategy could make a difference with random data - but I guess it boils down to the fact that some sequences of images contain slightly fewer or slightly more images by chance.

It might be worth doing a blog (or essay) devoted entirely to this effect and its significance or otherwise.

Do you think media types score extra well on presentiment? Perhaps you could persuade one of them to do a longer series for you!

I wonder if you could get some sort of feedback in your experiments - say an indicator that would show if if the previous image generated a curve in the right place. I know it would be noisy, but the brain is awfully good at discerning pattern in noise. If people could learn how to do presentiment....

Atheistic Mystic said...

I would love to see a compilation of demos from the documentary films and TV shows.

burning said...

Thought you might be interested in DoublePalm.com (http://fe19.news.sp1.yahoo.com/s/prweb/20080917/bs_prweb/prweb1336184).

Dean Radin said...

... but I guess it boils down to the fact that some sequences of images contain slightly fewer or slightly more images by chance.

An anticipatory bias can occur if a participant systematically becomes more anxious after each successive calm target until an emotional target appears, and then resets to a low anxiety condition. Computer models of such a strategy do show a bias that can mimic a presentiment result, if you use dichotomous targets, and if you run small numbers of trials in a single session, and if you don't pool trials across sessions.

The problem is that in the real world none of these assumptions are valid. I discussed this in the article mentioned at the mind-energy.net forum. Colleagues who've examined their own presentiment data confirm that realistic anticipatory models cannot account for the actual physiological results observed in these experiments.

If people could learn how to do presentiment....

This and other ideas haven't escaped the notice of entrepreneurs. A few years ago a colleague obtained a worldwide patent on the presentiment process, for use in intuition training applications.

David Bailey said...

Has the training idea worked out? I don't mean has it made money, but does it seem to have produced a real effect?

Have you tried any of the grauduates on your experiment?

Dean Radin said...

I haven't seen any actual applications yet based on the patent. But it's an indicator that someone thinks there's intellectual property worth protecting.

David Bailey said...

I suspect that some sportsmen may have developed a presentiment response. Think of how golfers or snooker players stare intensely at the ball after they have hit it - when logically it is too late.

Maybe if they see it do well or badly, that experience feeds back to the time before they hit the ball.

David Bailey said...

Dean,

Perhaps you could clarify something for me. When I look at a typical presentiment graph - such as those at the head of this blog, am I looking at the actual averaged data connected point to point, or have you fitted the data to a polynomial or something?

I ask, because the noise on these curves seems almost non-existent.

Dean Radin said...

This is the actual ensemble average, not a fitted curve. One of the nice things about working with SCL, as opposed to most other physiological measurements, is that the signal is nice and smooth.

blogtrotter said...

Dear Dr. Radin,

I believe the strength of your effect is indeed too great to be explained by any of the proposed artefacts. This sort of test is indeed worthwhile. I just finished Ross Targ & Jane Katra book "Miracles of Mind" - Ross is right to lament the retreat to the lab: the more natural macro- tests are better, and your natural reactions are a case in point. I also think more investigation of poltergeists should be done - e.g. that recent case in Catania di Coronia, Sicily, of a village evacuated by mysterious firs.

j said...

The curve is an averaged curve. If you compute an average curve, it is possible that future details affect earlier parts of the curve. Is it possible to show a non-averaged curve with raw data?

Dean Radin said...

The ensemble average shown here does not smooth within a trial (i.e., through time). It takes the mean across trials. Smoothing is allowable in presentiment analyses, but obviously the smoothing algorithm used cannot include future datapoints.

David Bailey said...

I wonder if you have tried any way to get the noise down - so that a useable prediction can be made before each event.

Do you think the noise would go down if you measured several different arousal indicators and combined them in some way, or does the noise simply reflect the way the person's arousal is varying - so all the measures would follow each other.

Presentiment seems frustratingly just below the S/N level at which dramatic demonstrations would be possible - such as scoring above chance in certain kinds of gambling!

Dean Radin said...

I've tried a number of ways of reducing the noise, including combining multiple physiological indicators. The most successful method I've found so far involves measurements based on the behavior of the eye. I'll mention more about this after the paper describing the experiment has been published.

Robin said...

Dr Radin,

Can you clarify? Are you saying that the pictures in the presentiment experiments did not fall into the categories upon which the analysis was based, ie "calm" and "emotional"?

Dean Radin said...

The analysis does use the categories of calm and emotional. But keep in mind that the pool of available pictures ranges from very calm to very emotional. So most of the pictures viewed will not be particularly calm or emotional, and thus those are the pictures excluded from the final analysis to ensure that we're only dealing with a strong contrast in affect. Is that any clearer?

Dean Radin said...

Update: Last week I ran another demo as part of an interview, and this one turned out significantly negative, meaning opposite to the standard prediction. In discussing this with the interviewer, he said that his reaction to a possible upcoming emotional event is to become extremely calm and quiet, rather than to become anxious. I had previously noticed that some people (mostly men) showed this same type of significantly reversed presentiment effect, but I hadn't thought to ask them how they typically deal with anticipated emotions or (putting it in more common terms) potential danger. In future tests I'll see if this factor predicts how they perform in this experiment.

David Bailey said...

Did the curve for the negative presentiment guy continue negative after the image was exposed - corresponding to his explanation?

Dean Radin said...

After the stimulus the average emotional curve went up a little as compared to the calm (not significantly). Nothing like the massively significant difference that is observed in people who show a positive presentiment result. It's rare to find someone who systematically shows a relaxation response after actually seeing highly emotional pictures.

Robin said...

Yes, thank you, that clarifies that my initial understanding or the picture selection was largely correct. But I am interested in why you think the gambler's fallacy only applies to dichotomous data. It seems that it applies in a number of situations - even random walk data like stock market graphs.

And it is not even certain that only the gambler's fallacy could produce this kind of artefact.

My simulation does not depend on dichotomous targets, only whether the prior stimulus has high or low emotional impact, so I don't know what difference variation in the emotionality of the pictures would have.

And why do you think simulations can't have large numbers of sessions or data pooled across sessions? I believe my simulations have pretty much the same structure as your experiments.

Robin said...

A further couple of points.

The anticipation effect requires no "reset" required as you say - I have no such reset in my simulation. All it requires is that there is a slight effect in the opposite direction during "emotional" trials.

Further I am surprised that you and your colleagues are so certain that the anticipation effect is not present in your data.

I can't even detect it in my simulation data without carefully targetted tests.

Tor said...

Dean,

About your article "Electrodermal Presentiments of Future Emotions", Journal of Scientific Exploration, Vol. 18, No. 2, I was wondering if you have any idea why two of the four experiments you conducted individualy turned out non-significant?

I can understand why the 4th experiment gave a non-signigifcant p-value, since it had much fewer trials. But experiment 2 had by far the largest amount of trials of all, but still turned up non-significant.

My impression has been that the presentiment effect (along with DMILS) was the most robust psi effect since so many have replicated it. Does it too suffer from this inherent unstability (critics would say unrepetability) that is common amongst psi studies?

Dean Radin said...

All studies involving human performance are permeated with noise, so not every experiment can realistically be expected to produce statistically significant results. That's why meta-analyses have become so widely used in the behavioral and medical sciences. They provide a way to assess whether there are repeatable patterns across experiments, without insisting for each individual experiment to be statistically significant.

Also, the particular type of presentiment experiment I've run uses pictures for stimuli, and people respond idiosyncratically to pictures, which adds noise to the experiment. My colleagues Ed May and James Spottiswoode have run a version using audio only, and obtained highly significant results (more so than any of my individual experiments).

I decided to use pictures because it makes the experiment more interesting for the participants, and it uses pictures that have already been employed in hundreds of mainstream psychophysiological studies.

Tor said...

Yes, the Ed May and James Spottiswoode study is intriguing.

It must be frustrating though at times to be dealing with such variable systems as human beings are. But that's reality, we are not dealing with a classical physics experiment here.

This brings me to another question.
Have you ever observed anything that seems to inhibit psi performance apart from boredom(or a rabid pseudo-skeptic yelling at the subjects)? Local sidereal time is a candidate to boost psi, but it seems to me that other factors must be at work in some studies too.
Or do you think all this variability can be explained by the subtle differences in the personalities etc. of the subjects?

David Bailey said...

If using sound yields more significant results, but the pictures hold subjects' interest better, why not combine the two?

Alternatively, why not hold their interest with a silent computer game of some sort, and blast them with noise as required!

Dean Radin said...

I once tried an experiment using both sound and pictures, but I personally found it too disturbing, and I don't ask anyone to participate in an experiment that I wouldn't want to do myself. I also tried a version that used a powerful electrical shock (Ed May built it out of a medical-grade shocking system), but that too was so disturbing that I didn't want to use it.

Dean Radin said...

... inhibit psi performance ...

Factors include: Effortful striving, anxiety, unfocused attention, poorly described instructions (for experimental tasks), lack of supportive set and setting, too much feedback (for some), not enough feedback (for others), annoying or confusing feedback, disbelief that the task can be accomplished (in the investigator or the participant), the personality trait of closedmindedness, a sense of being rushed, an inflexible target system (for PK), and so on.

In general, anything that would make you feel uncomfortable about performing at your best would also work against psi performance. There are probably some physical factors too, such as lunar phase or LST, but I think psychological and psychophysiological factors are stronger modulators of performance.

David Bailey said...

How easy is it to measure skin conductance reliably (e.g. can you get a device that plugs into a USB port!) - I mean the presentiment experiment sounds like something that could easily be set up in a school science lab.

If enough people tinker with your experiment, you never know what might emerge!

Dean Radin said...

The physiology monitors made by Wild Divine and Heartmath are affordable. Either of those systems could be used to create a presentiment experiment by someone adept in programming. Of course, as Rupert Sheldrake has shown, you really don't need fancy equipment to do a psi test. Rupert's "feeling of being stared at" or "telephone telepathy" experiments indicate that practically anyone, anywhere, who wishes to seriously conduct an experiment can do so at very little cost.

anonymous said...

Have professional athletes been tested for more sensitive or longer range presentiment effects? One could conjecture that if short term presentiment is naturally occuring in everyone then the best athletes in basketball, hockey, tennis, the martial arts, or batting and infielding in baseball, might have stronger presentiment effects which help explain their talent.

Dean Radin said...

> Have professional athletes been tested ...

Not to my knowledge, but it's a good idea. There are fewer than a handful of people in the world conducting presentiment experiments. By contrast [rant on] thousands of people are engaged in conventional psychophysiological research who have the equipment and skills to do similar studies, but prefer not to [rant off].

Nick Bentley said...

Above, a poster named Robin made some interesting methodological criticisms related to the Gambler's Fallacy to which I am not able to find your response.

Like Robin, I can't figure out why dichotomous stimuli are necessary, or why a reset is necessary.

BobH said...

Dean

In "Entangled Minds" (PP.165) you say "After the button press the computer waits 5 seconds, selects a picture at random...[and then] displays it on the screen for 3 seconds".
In this blog you say "...the moment the target is selected and shown, at second 0...".
I understand these quotes to mean that when the presentiment effect starts, the picture has not yet been selected by the computer program.
This implies that we can sense something that hasn't happened yet.

If the situation was that the computer had already selected the image at second -3, the implication could be that we can sense information that is being witheld from us (ie. the computer already "knows" the image it will display) - which is still amazing (akin to knowing that we are being started at, remote viewing and telepathy) but is different from sensing "the future".

I was wondering if the 5 second waiting period was a true null period of pure waiting or if the computer made use of this time to execute the random algorithm and had possibly already selected the image by the time the presentiment effect starts?

Dean Radin said...

> I was wondering if the 5 second waiting period was a true null period ...

It is.

One or two tests of this type that I did when I first started to explore this paradigm had the target pre-selected, but hidden, during the presentiment period. All of my later studies, and those of most of the replications, are genuinely precognitive.

Spirits1234 said...

Hello Pr. Radin!
I know it's an old post but could you please spare some minutes to reply to my question, I would be really grateful for that.

I'm doing a class presentation on the Presentiment Experiment you described.
However while browsing for materials on the net I came across a skeptic website "http://forums(dot)randi(dot)org/showthread(dot)php?t=123007", which might be the most annoying website for parapsychology researchers.

However Robin mention a good counter argument to your argument about the statistical analysis method used.

I'm a bit confuse about what to really believe, Could you please reply provide a response on your side?

I really appreciate the efforts you've done on Parapsychology research, I really hope for the success of parapsychology.
But I hate even more to have doubts in what I believed to be the most prominent research in Parapsychology.

Thank you.
Allen.

Dean Radin said...

What is more credible? Comments left on a skeptics forum or blog, or articles that have been published in peer-reviewed scientific journals? Peer review is important because the reviewers thoroughly know the literature, so they can evaluate ideas in an informed fashion. People proposing ideas in forums often come up with clever explanations, but they are rarely aware that their ideas have already been thoroughly discussed.

This is a good case in point. Those of us who have conducted these experiments are well aware of the proposed anticipatory explanation. It's a form of gambler's fallacy, and my colleagues and I have specifically tested this idea in the data collected in our experiments. The bottom line is that the proposed explanation is not supported by the actual data.

It is not the case that skin conductance (or any other physiological measure that has been studied to date in these experiments) is systematically influenced by the previous target. What we see in these studies is a sudden rise in pre-response anticipation just before the emotional target.

Spirits1234 said...

Wow!
Thank you professor!
I wasn't expecting such a quick reply.
I will take your comment into consideration and I will try to make my presentation better.
Thank you for your quick reply.

I also asked this question on the Skeptiko website, am I allowed to post your reply?