readings> psi - remote staring controversy

Does the mind have paranormal powers? One of the most recent battlegrounds has been the remote staring experiments carried out by sceptic Richard Wiseman and believer Marilyn Schlitz. Guess what results each gets?

IF YOU have ever had that creepy sensation of being stared at from behind, then you know what some of the subjects in Richard Wiseman's lab are feeling right now. Sitting in a soundproofed room with their skin wired up to an arousal monitor, they know that for some of the time someone may be staring at them via a CCTV link. The question is: are the subjects subconsciously aware of when they are being watched?

Wiseman, a psychologist at the University of Hertfordshire in Hatfield, UK, doesn't think they do. He's been doing experiments like this for years and he always gets nish, nada. But his co-worker, parapsychologist Marilyn Schlitz of the Institute of Noetic Sciences in Petaluma, California, begs to differ. When she does the experiment, she usually gets a slight, but significant, positive result. And so the two have set out to resolve the discrepancy in a head-to-head showdown.

It sounds like just another investigation into the paranormal, but the Wiseman-Schlitz study represents something of a watershed. The sceptic-versus-believer experiment is designed to nail down one of parapsychology's longest-standing controversies, the "experimenter effect". This is the curious phenomenon whereby the outcome of an experiment hinges upon the beliefs of the person running it. Believers tend to get positive results. Sceptics don't.

The results will be available later in the year. But arguably it's not so much the outcome that matters as the fact that the study is taking place at all. The experimenter effect is the skeleton in parapsychology's cupboard. Parapsychologists once saw nothing wrong in using it as an excuse for ignoring inconvenient experimental results. Some even argued that the fact some experimenters got results and others didn't was proof that psychic powers (or "psi") existed.

But about 20 years ago, parapsychology researchers agreed to stop invoking the experimenter effect, aware that it was undermining their attempts to prove the existence of the paranormal. It simply made it too easy to dismiss studies that didn't get results, and concentrate on the ones that did.

Yet parapsychologists are increasingly bringing the experimenter effect back into the debate. The inescapable fact, they say, is that it is there in the data. Whatever kind of experiment you run, whether into extrasensory perception or telekinesis, believers get results and sceptics don't. "In my opinion, the strongest predictor of ESP results generally is the identity of the experimenter," wrote parapsychologist John Palmer of the Rhine Research Center in Durham, North Carolina, in a recent special issue of The Journal of Consciousness Studies (vol 10, p 57).

The experimenter effect itself is now an object of intense research. As well as the flagship study of Wiseman and Schlitz, almost every other parapsychology researcher is conducting some kind of test. There are also new explanations of the experimenter effect on the table. Some parapsychologists claim that it arises not through the experimenter's influence over mind or matter, but because they use their extrasensory powers to pick the right moments to sample a fluctuating process and catch any fluky, but natural, departures from randomness.

The return of the experimenter effect is leading some sceptics to claim that parapsychology has been unable to produce convincing data and so is regressing back into the pseudoscientific dark ages. But supporters of parapsychology turn that reasoning on its head: there is an experimenter effect, they say, and the imperative is now to explain it. So after a period of uneasy accord during which the two sides have worked together under mutually acceptable rules, a new schism is developing.

Until about 20 years ago, the experimenter effect was an integral part of parapsychology, along with a number of other slippery concepts that seemed to capture the weak and erratic nature of the phenomena. A run of good results seemed to be followed by a run of bad, almost as if the odds wanted to right themselves, an effect parapsychologists dubbed "psi decline".

Then there was what is known as "psi missing", where post hoc analysis of results revealed that the subject kept missing the target by a slender margin. For example, in an ESP test in which the subject had to guess the suit of the card that the experimenter was looking at, they would consistently say "heart" on the turn just before or just after a heart turned up. And then finally there was the experimenter effect - the feeling that belief matters in getting positive outcomes. Some parapsychologists even complained that having a doubter visiting the building seemed to spoil their day's work.

In the 1970s, parapsychologists such as Adrian Parker, now at the University of Gothenburg in Sweden, began running studies to see if the experimenter effect might be genuine. For example, he recruited students as experimenters and then briefed them to expect either positive or negative results from a test. "It seemed completely logical to us that if psychic powers existed, then there must be the possibility of an experimenter effect. So it was obviously something we needed to get a handle on," Parker says.

But in the 1980s the field changed tack. After years of rancour in which sceptics and believers stood on opposite sides of a divide, it was agreed that the only way forward was to establish some mutually acceptable standards. Sceptics agreed to ditch their presumption of artefact or fraud whenever a positive result came in; in return, parapsychologists agreed to drop all forms of special pleading, and concentrate on getting replicable results. No single experiment could be trumpeted as proving the existence of psychic powers. And the same experiment would have to keep producing results no matter who was running the show - sceptic or believer.

This, of course, meant sacrificing the experimenter effect. Parapsychologists had to rely on the fact that if the experimental designs were sufficiently robust and enough different experimenters were involved in their replication, they wouldn't need the experimenter effect to get positive results. The presence of doubters might water them down, but should not eliminate them.

Parapsychologists took other healthy steps. Led by the example of the Koestler Parapsychology Unit at the University of Edinburgh, UK, there was a broadening of the remit to take in the "psychology" of parapsychology in general - normal, rational psychological explanations for why people might believe in the paranormal, or be fooled into believing in it. In this way, parapsychologists could come to be their own best critics, able to separate what might be true from what was mere wishful thinking.

Psychologist Matthew Smith at Liverpool Hope University College in Liverpool, UK, says this mixed approach has made parapsychology more academically acceptable in the UK and Europe over the past decade, and a routine part of many psychology departments.

"It has become a popular option for third-year students," Smith says. "Because issues like statistical artefact are such a concern, they can learn a lot about what is needed to do science rigorously."

So what has all this rigour done for the status of parapsychology? As usual there is no consensus. Believers claim the results show there is an effect, sceptics say they don't. The major battles have revolved around meta-analyses that blend all the studies.

Wiseman and Julie Milton, then of the University of Edinburgh, did one on an ESP technique called ganzfeld that showed no significance for some 30 combined studies, run by both believers and sceptics (Psychological Bulletin, vol 125, p 387). But the believers came back at them, saying they omitted a couple of recent, highly significant studies they knew about but left out by setting the cut-off date conveniently just before them. So it is now a matter of whose meta-analyses you want to believe.

What everyone agrees, however, is that there is an experimenter effect in the pattern of results. Believers are somehow making their experiments work - the question is how.
This is where the Wiseman-Schlitz experiment comes in. While they are not happy to see it portrayed as a battle between sceptic and believer - Wiseman says they are merely two experimenters with different track records - others see it as a test case for the idea that belief alters results.

The experiment itself is simple. Subjects sit in an isolated room hooked up to electrodes which measure their arousal levels through slight changes in the sweatiness of their hands. In another room, the experimenter can see the subject on a CCTV. The experimenter's job is either to stare at the subject, or look away, according to a random 16-minute schedule divided into 30-second blocks. The hypothesis is that if the subjects know they are being stared at, they should show a detectable shift in physiological arousal during the times they are under surveillance.

Schlitz and other parapsychologists have been claiming small but statistically significant results from this experiment for more than a decade. In the mid-1990s, Wiseman - who was a professional magician before becoming a psychologist - thought he would have a crack at the same experiment. He says the set-up was attractively simple and the claimed results sounded clear-cut. But when he ran the experiment in his own lab, he got nothing.

The story might have ended there had Wiseman and Schlitz not met and found they liked each other enough - and trusted each other enough - to carry out a joint study. And so for the past eight years, Wiseman the sceptic and Schlitz the believer have been doing the remote-staring experiment while swapping labs and sharing subject pools.

"Richard probably wanted to find the ways I was cheating or whatever," Schlitz laughs. "But I think he was genuinely curious also."

First Schlitz came to Wiseman's lab in Hertfordshire. Each tested 16 subjects, mostly undergraduate psychology students. Wiseman got results that were flat chance; Schlitz reported a positive outcome, though these were on "the very knife-edge of significance", as Wiseman puts it. Then Wiseman flew over to Schlitz's lab in Petaluma to repeat the experiment, this time with subjects who were using the institute as a New Age retreat. The result was the same - nothing for Wiseman and knife-edge significance for Schlitz.

Which brings us to the present day. It was clear the study ought to be repeated with a much larger number of subjects so as to give it real statistical power. But just as importantly, Wiseman and Schlitz needed to separate two possible sources of any putative experimenter effect. Does it come from the experimenter's psychic ability, or is there another explanation?

As even the proponents of psychic powers point out, non-psi factors could bias the running of the experiment. Faced with a cold and sceptical experimenter, apparently doubting that the study is going to produce any results, perhaps subjects are inhibited in some way. Conversely, a "psi-conducive" experimenter could be one who conveys the necessary warmth and belief to allow shy psychic abilities to flower in a laboratory setting.

"These experiments can be long and tedious, so it's quite plausible that even if I tried to conceal my feelings, my boredom might still show through and affect the attitudes of subjects," Wiseman admits. "Marilyn is naturally much more enthusiastic and positive in her interactions with the subjects."

So in the latest study, due for completion this spring, the duo are splitting the "meet-and-greet" from the experimental phase. On some trials one will do both the introductions and the remote staring. On others, they will divide the jobs between them. As a further measure, the interactions between experimenters and subjects will be video-recorded, then later independently rated for factors such as warmth. With this careful crossover design, and more than 100 subjects involved, the source of any experimenter effect should become clearer.

Of course, the two possible sources of the experimenter effect rely on the existence of paranormal powers. But it's hard to know where else to turn: even Wiseman accepts that there is an unexplained experimenter effect in the results to date. Naturally the sceptical camp is hoping the new study simply finds no sign of an experimenter effect. Or perhaps that some statistical artefact will be unearthed to explain it.

Another "out" for sceptics is always going to be fraud. Some minor concerns about the statistical analysis and security precautions in the earlier Wiseman-Schlitz studies have been addressed in their latest experiment. However, Wiseman stresses that none of the concerns were major issues.

"It is a basic misunderstanding that I'm here as the sceptic to be the policeman watching Marilyn," he says. "It doesn't work like that. We take reasonable precautions in the experiment, but nothing extraordinary." Wiseman says the answer to all concerns must, as always, lie in replication. "If it is only us that gets this weird result, then it is uninteresting," he says. "Other pairs of sceptics and believers would have to work together and find the same thing happening before we could get too excited."

Whatever the outcome of the experiment, the resurgence of the experimenter effect is contributing to a new schism in parapsychology. On one side are those like Wiseman who feel that what has been proved is that psychic powers do not exist - or more accurately that, because you can't prove a negative, the null hypothesis still holds. He says every time a new kind of experimental design comes along, at first it gets good results, invigorating the field, and then these results fade away.

Wiseman says either the procedures get tightened up, ruling out artefacts, or perhaps the first results are just an honest statistical fluke that cannot be repeated - a kind of "file drawer" effect in which experimenters have a go at a great many different parapsychology tests, but only report the ones where chance guessing seems to produce a significant result.

But on the other side are those like Parker who argue that psychic powers are real and the experimenter effect is part of them. Parker feels it may well be time to leave the doubters behind, take the existence of psychic powers as sufficiently proved and turn scientific attention to the finer details of how they work.

So, as usual, there are no black-and-white answers in prospect. In many ways, it is the sceptical community that is on the back foot, unable to explain away the results in terms of cheating, artefact or fluke. They are back to making suspicious noises about why believers get results. But there are downsides for the parapsychologists, too. By turning the spotlight back on the individual experimenter, they are again inviting the old accusations of methodological incompetence and fraud.

Ultimately, the real loser is the delicate accord that was struck some 20 years ago, when sceptics and parapsychologists agreed how the field should be run. That agreement at least gave us some hope of a definitive answer. Without it, one way or the other, the world again becomes divided by a question of belief.

psi research and its results

Paranormal powers can be divided into two types. One is extrasensory perception (ESP, also known as telepathy), which is the ability of two minds to communicate through an unknown channel. The other is telekinesis, the ability of minds to affect matter. Attempts to prove that these powers exist usually use one of three basic experimental designs.

MICRO-PK is a test of telekinesis - PK stands for psychokinesis. Early studies focused on moving large physical objects, such as a suspended metal strip. Most experiments now concentrate on microscopic events such as the random white noise produced by a diode transistor, or the diffraction patterns of beams of photons. Subjects view the output of the particular random event generator on a computer terminal and try to "bend" the flow of events away from chance.

Results: Many experimenters fail to show an effect. Those that do typically report only the smallest deviations from chance - but because millions of events are involved, the results are claimed to be statistically significant.

GANZFELD - German for "whole field" - is a test of ESP. A subject sits in a state of isolated sensory deprivation, eyes covered, white noise playing in headphones, in the hope of maximising sensitivity to any psychic signal. In another room, a sender views one of four randomly chosen video clips played repeatedly for 30 minutes. Afterwards, the subject views all four possible clips and picks the one closest to any mental imagery experienced during sensory deprivation.

Results: The subject has a 1-in-4 chance of guessing right, so should average a "hit rate" of 25 per cent. The latest meta-analysis of 40 studies claims an overall 30 per cent hit rate - a significant result. But other meta-analyses conclude that the results are within the bounds of chance.

DIRECT MENTAL INTERACTION WITH LIVING SYSTEMS (DMILS) investigates both ESP and telekinesis. It is a general term for attempts to use a mental connection to influence a distant biological system. This could be causing changes in the physiology of another person, as in remote-staring experiments, or affecting the growth rates of seedlings and yeast cultures. Statistics are used to contrast effort versus no-effort trials. Proof would be taken as evidence for psychic healing.

Results: Latest survey of 40 remote-staring studies reports a small, yet significant deviation from chance. But when only the seven "highest quality" studies are considered, the positive effect disappears.

home> back to readings