Parapsychology Journalism: The People, The Theory, The Science, The Skeptics
Because of the controversy surrounding psi and its existence, it’s vitally important to read the source material. It is not enough to examine the conclusions of any but the most simple experiments. We have to ask the question: What experiments actually took place? How were they conducted? What controls were used? How was the study safeguarded from cheating? Unless we know these things, we cannot make good judgments about what took place.
In no other study of psi is this more important than when we examine mediums: people who get information from the dead. Mediumship has a long and sordid history of fraud, so scientifically testing it must necessarily demand excellent safeguards against cheating. A poorly conceived and executed study would ruin the reputations of everyone involved.
The experiments I am examining were conducted by Gary Schwartz, PhD as outlined in his book: The Afterlife Experiments: Breakthrough Scientific Evidence of Life after Death. I will also be referring to an article by Ray Hyman, PhD, who presented a skeptical counter argument to this book.
The HBO Dream Team Experiment: This was a multimedium/multisitter research experiment. The mediums included famous psychics: George Anderson, Rev. Anne Gehman, John Edward, Suzane Northrop and a local medium, Laurie Campbell. And two sitters. One was obtained by HBO and the other was found by Dr. Schwartz and his staff. The requirement was that the sitters have six deceased people that they were close to.
What is striking about this experiment is that the famous mediums would participate at all. If you are using cold reading and trickery, as skeptics presume here, no way do you want to lose control of your environment. You are basically gambling that you can outsmart the researchers. For John Edward, a TV personality and the subject of scathingly personal skeptical attacks, this could easily amount to professional suicide.
You run a high risk that people doing your detective work (to discover facts about the sitters) will be eventually discovered; you run a risk that the researchers are smarter than you think and have devised experimental methods you cannot defeat; you run a risk that someone will see something odd and start talking and most of all, you will have people not under your control who are attempting to anticipate and defeat your every move. These risks so completely outweigh any possible gains in credibility and status one could gain from the event that participating would be an insane choice for a cold or hot reading entertainer posing as a psychic.
Not one, but four high profile mediums took part in the experiment. It is extremely difficult to imagine them participating unless they could actually do as they claimed. The only catch in the experiment was that the sitters vouched for themselves that they had no previous contact with the mediums or anyone associated with them. They wrote down the names and details of the deceased people who would be involved and this document was stored with no one else knowing the contents, not even the researchers.
To prevent visual cues, the sitters were put behind opaque screens; the mediums would not be able to see them during sessions and the sitters would only be able to give short yes, no type answers. Afterwards, the sitters were required to go through the transcripts and rate the answers: +3 for a hit, +2 for a probable hit, +1 for a possible hit down to -3 for a definite miss. The result for the mediums was an 83% success rate. A control group of 68 U of Arizona students who were just guessing came up with 36%. So unless the mediums came up with an incredible deception that allowed them to get correct information on someone that they couldn’t even see, they beat chance by a significant amount.
The Russek Protocol: This experiment consisted of two parts: A silent sitter period where no verbal contact with the sitter would occur and then a period during which they would be allowed one word yes/no responses. There would also be five sitters, not just two. Thus, particularly during the silent period, there would be no way for the medium to know who was in front of them. Knowledge obtained through detectives would not work. This was put into place for the following experiment.
The Miraval Experiment: The mediums were the same except that George Anderson did not participate in this one. The results were 77% for the silent period and back up to 85% for the questioning period. It seems rather unlikely that a cold reader would participate in an experiment where they could not see or hear the person they were asked to provide information about.
The Canyon Ranch Experiment : In this experiment, the requirements were even more stringent. To prevent the assistance of conspirators, floor to ceiling sheets were used between the mediums and the sitters. During the first ten minutes, the sitter would not speak, and later, someone else would speak for the sitter. The medium would have no clues about the sitter whatsoever. There were five sitters. They would be sequestered in a room where they could be neither seen nor heard by the mediums. As usual, the sitters wrote down the pertinent information beforehand and this was sealed away to be opened after the experiment ended. The sitters were also asked to score not only their own reading, but everyone else’s as though they were their own to be sure that vague information that fit everyone was not the cause of the success rate.
The results were that readings for a specific individual got twice as many +3 hits as when they were doing the scoring from someone else’s reading. So the actual medium readings were twice as good as the controls. The hit rate went down in this experiment, averaging about 40%. But one medium drew a complete blank with one of the sitters and this contributed to the result. It did demonstrate that the mediums could significantly beat chance even with this added hurdle.
The Campbell White Crow Readings: Medium Laurie Campbell participated in an experiment in which she did not know the three sitters and her only contact with them was over the phone. Further, for half an hour before initial contact, she would write down her impression of the sitter. Then the person was dialed up and their phone was placed on mute so that they could not hear the medium for ten minutes of the reading. After the completion of the this, the phone was taken off mute and the medium shared her impressions with the sitter; asking for confirmation or denial of various things she had said. The accuracy ranged from 80% to 100% per deceased person. In a double blind version of this experiment in which not even the experimenters knew the order of the sitters, the medium obtained 60% accuracy. When the sitters scored their readings, the hits that the medium got were well above the control.
As you can see, the experiments were ever more stringent and the mediums still performed better than the controls. One thing that must be taken into account here is the complexity of cheating in this class of experiment. The elements for successful cheating are 1. Knowing who the sitter is and 2. Obtaining accurate information about that sitter that can be used during the session. How hard would that have been to accomplish? The answer of course, is extremely hard. With all these sitters and all these mediums, the risk of failure would have been very high indeed. The only reasonable conclusion is that this was genuine mediumship.
What do the skeptics say?
The main piece of literature counter arguing these studies is the article cited above by Ray Hyman. It was published in the Skeptical Inquirer, a non peer reviewed skeptical publication. Victor Zammit addresses this article in his counter counter argument in his usual abrasive style in this article on his website. (I’m assuming his website is not peer reviewed either.) Gary Schwartz provides his own comprehensive review of Ray Hyman’s critique.
As someone who is skeptical of skeptics this article has some issues which make it unconvincing to me. The following represent my opinions:
1. The Big ‘Ol List of Defects: Ray Hyman lists 11 areas where he finds defects in these experiments, but does not elaborate on any of them nor does he assign them any varying levels of importance in the outcome of the research. If the defects were obvious to anyone who had read both the book and the article, then perhaps they would need no clarification. For example:
1. Inappropriate control comparisons
2. Inadequate precautions against fraud and sensory leakage
3. Reliance on non-standardized, untested dependent variables
4. Failure to use double-blind procedures
5. Inadequate “blinding” even in what he calls “single blind” experiments
This stuff needs explaining and this criticism must be evaluated for its importance to the experiments. In the absence of clarification, these arguments come across as nothing more than heavy handed nonsense. (“You must be wrong because I have ELEVEN arguments to use against you.”)
2. Ignoring the big picture and focusing on details: The arguments that Dr. Hyman makes are long, boring and technical which isn’t normally a problem, but in this case the whole tangled ball of yarn can be cut with one simple statement: “Could a cold reader have done this?” If the experiments had adequate safeguards against deception, then the rest is just quibbling.
3. Armchair criticism: Dr. Schwartz challenges skeptics to prove that cold reading could have produced the same results under the same conditions. The skeptical counterpoint that Dr. Hyman makes is this:
The skeptics and the scientific community have no responsibility to show anything until you provide them with data collected according to well-established and acceptable standards. The responsibility is yours to first provide us with evidence for your hypothesis of survival of consciousness that is gathered according to the appropriate scientific standards which include controlling for sensory leakage; devising dependent variables that are relevant, reliable, and valid; and using control comparisons that are meaningful.
In other words: “No, we’re not going to do anything that might prove us wrong.” This approach, by Dr. Hyman, leaves the skeptics a loophole they could drive a truck through. It is unprofessionally vague, allowing the criticism to be endless. What sensory leakage? What dependent variables? What control comparisons? How do these terms translate into testing procedures that skeptics will accept? Dr. Schwartz is not himself a magician and cold reader. He cannot be expected to anticipate all their tricks and skills unless someone comes forward to show him.
Magicians have always come forward in the past to perform this function for parapsychologists and indeed, Dr. Schwartz enlisted the help of one in designing his experiments. If there is something about cold reading he doesn’t know, someone has to show him so he can improve his testing. Dr. Hyman is claiming that the testing was faulty without concretely showing how it was faulty. He has abdicated the most important duty of a skeptic.
4. No end in sight: A skeptical review of any experiment is useless if it does not include a concrete list of suggestions, that if met, would satisfy the objections of the skeptic, who would then abide by the results of the revised experiment(s). The way the thing is written it appears to me that Dr. Hyman was deliberately avoiding the possibility of confronting a successful experiment.
It is for these reasons that I don’t find Dr. Hyman’s article convincing. However, if a clear and understandable argument was made that the experiments were flawed. And if this could reasonably be corrected by further experiments, which, if successful would require the involved skeptics to fully concede, then I would take notice.