Ganzfeld Experiments

Today we're going to enter a quiet, darkened room, sit comfortably, and prepare to receive psychic imagery, in what's often claimed to be the most convincing evidence for the reality of psi — psychic abilities. The idea of being able to transmit thoughts from one person to another is so compelling that there's never been a shortage of researchers hoping to find a way to develop it. We all wish we could have such a superpower, so we all want this to be true. Today's subject is ganzfeld experiments. Ganzfeld is German for "whole field", referring to its method of replacing the whole of your field of perception. Let's take a close look and see what it is, how it works, and — most importantly — whether it does indeed promise to be proof of psi.

A ganzfeld state is a bit different from sensory deprivation, as made famous in the movie Altered States. In sensory deprivation, the idea is to remove all stimuli, audio, visual, thermal, and tactile. Ideally the subject is placed in an isolation tank, a coffin-like device in which you float in a dense saline solution, the temperature is a constant, comfortable ambient temperature, and it's completely dark and quiet. You see, hear, and feel nothing. Sensory deprivation has often been used recreationally, both with and without hallucinogenic drugs, for its ability to make the imagination seem surprisingly real, given the lack of competing stimuli.

However, in ganzfeld, the idea is to instead provide homogenous stimuli. The subject, called the "receiver", sits comfortably in a recliner, wearing headphones playing gentle white noise. The room is bathed in red light and the receiver wears translucent cups over the eyes, so all they see is a uniform, featureless red. They are relaxed and cozy. That's the physical setting of the experiment. Two other people are involved: an experimenter and a "sender". The sender, in an isolated room where they cannot be seen or heard by the receiver, concentrates for 30 minutes on a "target", which is some object or video clip or something. Throughout the 30 minutes, the receiver is supposed to verbally recite what they see or imagine. The experimenter, who is also supposed to be isolated from both the sender and the receiver, records what the receiver says, and usually keeps notes about what they describe.

At the end of the 30 minutes, the receiver is shown the actual target upon which the sender was focusing, presented alongside with three other control objects. The receiver guesses which of the four most closely resembles their impressions during the ganzfeld session. Pure chance predicts a 25% hit rate. But ganzfeld experiments became famous within the parapsychology community because experimenters consistently found a significantly higher hit rate; closer to 35%.

The history of ganzfeld experimentation is essentially the history of a particular battle between skeptics and believers; a cordial battle, but a battle nevertheless. Beginning in the 1970s, the leading proponent was American parapsychologist Charles Honorton, a staunch believer in psychic abilities, who was dedicated to finding a reliable scientific method of establishing the reality of psi. Honorton's idea was that whatever psi abilities many people may have is lost in the sea of constant stimuli that we're all receiving all day long. We see, we hear, we touch, we think, to such a degree that if we did receive a psychic impression we'd never recognize it as such. So by placing subjects into a ganzfeld state, it's thought that the signal-to-noise ratio would be increased, by shutting off all that noise, and subjects might be more likely to recognize a psychic transmission.

Across the line of battle was Ray Hyman, at the time a professor of psychology at Harvard. In the 1980s he came across Honorton's body of work, said to be the best evidence yet for psi. Hyman studied it carefully, and came away unconvinced. In his assessment, the positive results so flaunted by the parapsychologists was due to methodological error. In 1985, Hyman published an article in the Journal of Parapsychology called "The Ganzfeld Psi Experiment: A Critical Appraisal".

Unimpressed right back, Honorton published — in that very same issue of the journal — "Meta-Analysis of Psi Ganzfeld Research: A Response to Hyman". Clearly, there was a difference of opinion.

Before we look at what happened next, let's hear out both Honorton and Hyman to see exactly what was right or wrong with the research. Most of this involved what are called meta analyses, which is when you combine the results of multiple studies with the goal of getting a better idea of what the whole body of research in a field has concluded. Meta analyses are tricky animals, because studies can be conducted in so many different ways, and are often of greatly varying quality. All sorts of statistical methods can be employed (rightly or wrongly) to try and account for and control these differences. It's not surprising — in fact it's to be expected — that researchers can come up with greatly differing findings doing meta analyses on the same set of studies.

Hyman brought a considerable amount of skepticism to the table, so I was expecting his article to find all sorts of problems with Honorton's work, which covered 42 studies in which 55% showed positive evidence of psi. Hyman did report problems, however his own corrected analysis found not the random chance result of 25%, but a still-significant 30%, which, in itself, appears to still represent pretty amazing evidence that Honorton's receiving subjects were in fact receiving some kind of impressions from the senders. However, not so fast. The criticisms that Hyman found were inadequate randomization; sensory leakage (meaning that in some cases, the receivers could actually hear what was going on in the sender's room next door; in others, it was possible for things like the sender's fingerprints to be visible on the target object for the receiver to see); and inappropriate statistical analysis.

Mainly, Hyman felt that Honorton's work suffered from a type of statistical complication called multiple testing. In a nutshell, multiple testing is when you take more and more variables into account between two groups; sooner or later you're going to find more and more differences between them. These variables included the different ways that researchers had categorized the senders and receivers, cross referencing them to the results. They found that subjects were more likely to have positive results if they had been educated in a creative field; if they already had a strong belief in psychic powers; if they were extroverted; and if the experiment was conducted in a warm and welcoming atmosphere. Hyman believed that the positive results reported by Honorton were due, at least in part, to multiple testing effects that inappropriately considered these types of variables. Hyman also found that the "file drawer effect" came into play, which is when studies are abandoned when they end up not showing any interesting results. Thus, the body of published work was inappropriately skewed to include those results which showed a positive result, which is going to happen sometimes simply due to random variances. Hyman figured that, working backwards and accounting for the degrees to which various weaknesses were present in each of the studies, the actual size of the effect was zero. His closing line was:

It is concluded that this data base is too weak to support any assertions about the existence of psi.

Honorton's reply in the journal was in kind. He acknowledged all of these potential weaknesses, but explained how he had accounted for them, and still insisted that the results supported the existence of psi. This whole discussion got about as deep into statistics as anyone might reasonably (or unreasonably) want to go; but the net result is that the men had a disagreement on the analysis of the existing body of work. So now let's look at what happened next.

It was quite refreshing, and something that I wish I've seen more often. Honorton and Hyman got together and collaborated on a new article, hoping to find an analysis they could agree on. It was published in the same journal in 1986, entitled "A Joint Communique: The Psi Ganzfeld Controversy". Essentially, Honorton and Hyman agreed on the methodological weaknesses and on ways to fix them, but were not able to come to a consensus on the proper analysis of the existing studies. They concluded:

We continue to differ over the degree to which the effect constitutes evidence for psi, but we agree that the final verdict awaits the outcome of future experiments conducted by a broader range of investigators and according to more stringent standards.

The result of this collaboration was a process called the autoganzfeld study. This was a computer controlled version of the ganzfeld experiment, where randomization and the other methodological weaknesses identified by Hyman and others were eliminated. In expressing his satisfaction with the potential of the autoganzfeld process, Hyman wrote:

Honorton's experiments have produced intriguing results. [If] independent laboratories can produce similar results with the same relationships and with the same attention to rigorous methodology, then parapsychology may indeed have finally captured its elusive quarry.

So with replication as the goal, many researchers at many locations took up the torch and began performing autoganzfeld tests. Publications continued to be produced, many of them continuing to find positive — if small — results. It was this decade and a half of replication efforts that led to the next major publication on ganzfeld experiments. Honorton and his collaborator, Daryl Bem, published this time in a more mainstream journal, Psychological Bulletin, in 1994 (after Honorton's death). Their conclusion was optimistic, but measured:

The autoganzfeld studies by themselves cannot satisfy the requirement that replications be conducted by a "broader range of investigators." Accordingly, we hope that the findings reported here will be sufficiently provocative to prompt others to try replicating the psi ganzfeld effect.

Their article failed to convince. In 1999 the Bulletin published a thorough critique of Honorton and Bem's paper. Its authors were experimental psychologists Richard Wiseman and Julie Milton, and it focused on the failure of the popularly reported positive results to be replicated by independent researchers. After a deep discussion of all the problems found with Honorton's body of work, Wiseman and Milton concluded:

The new ganzfeld studies show a near-zero effect size and a statistically nonsignificant overall cumulation... The autoganzfeld results have not been replicated by a "broader range of researchers." The ganzfeld paradigm cannot at present be seen as constituting strong evidence for psychic functioning.

With the death of Charles Honorton in 1992, interest in ganzfeld has declined somewhat, though psi researchers such as Dean Radin have continued to support it. The best lesson to be learned from ganzfeld experimentation is not so much that the technique has failed as evidence for psi, but rather that it is indeed possible for skeptics and believers to work together in a productive, positive, and collaborative way to find the truth. Hyman and Honorton showed us that the mainstream and the fringe need not always be squared off with an us-vs.-them mentality, and reminded us that the best researchers, whether they're right or wrong, embrace their critics and work with them to improve the state of our knowledge.

By Brian Dunning
Follow @BrianDunning