Mediumship research began back in the late 1800s. Séances were popular and there were many mediums, both real and fraudulent, cashing in on the public interest. While most of the public believed in the supernatural, many scientists were skeptical and the best of them suspected that fraud was mixed in with some real abilities. This led to the formation of the Society for Psychical Research in 1882. It was intended as a specialized organization dedicated to serious investigation into not just mediumship, but hauntings, poltergeists, telepathy and other paranormal claims.
There were originally many spiritualists on board, in addition to more traditional scientists, which led to a clash of guiding philosophies. The spiritualists felt that the organization should operate from the premise that psychic phenomena were real and that they should be studied more in depth, rather than focusing on proving their existence. They lost out to the more skeptical faction and many of them left the organization.
Early Mediumship Research
The mediumship research was investigative. Researchers would attend séances, control for as much as they could, and observe the mediums in action. There were two types of mediumship: physical and mental. Physical mediumship was far more splashy. Horns sounded, tables rose, knocking was heard, and other phenomena were observed. This was also where the majority of fraud occurred; it’s a problem that still plagues these investigations to this day.
Nevertheless, two mediums rose above the rest. Eusapia Palladino and Daniel Dunglas Home. While Palladino would cheat at every opportunity, she would also produce convincing, verifiable phenomena for skeptical researchers when it was clear that cheating wasn’t an option. She was perhaps the most studied medium of her time. And because of her “uncouth” manner and tendency to cheat, she was also very controversial.
Daniel Dunglas Home by contrast, also produced remarkable phenomena, but unlike Palladino, was never caught cheating despite being intensively studied. (There is no first hand account of anyone ever witnessing his attempting to cheat.) There were, of course, allegations of fraud, but they were unsubstantiated.
Harry Houdini: Fraud Buster
Mediums, of course, had their skeptics, the most well known of which was Harry Houdini, the famous magician and escape artist. At first a believer in the afterlife and mediumship in general, he discovered how easy it was to deceive people by posing as a medium himself and employing what was later called the Barnum Effect. (Using generalized information that people take to be specifically about them.)
No one disputes that Houdini uncovered some fraud during his time as a debunker, but there were signs towards the end, in a spectacularly public clash with medium Mina Crandon and the magazine Scientific American, that he had gone off the deep end. He was accused of deliberately attempting to frame Crandon for fraud, argued with everybody and generally made a distasteful nuisance of himself. This is typical of pseudo-skepticism, where a person makes a claim to be an impartial observer and then demonstrates themselves to be so biased against the original claimant that no amount of evidence is convincing to them.
You can generally spot them by their 100% faith in their own opinions, lack of gray areas and absolutely 0% respect for anyone who disagrees with them.
Scientific inquiries into mediumship research in fact, have always been hampered by this type of critic that imagines flaws and/or fraud in everything they see. More on this later.
The Difficulty of Testing Mediums
Mediumship research, even within parapsychology, has received little attention over the past decades and only recently has interest begun to increase as a result of relatively recent research. One of the problems with mediumship studies is that there are possibilities for biases all over the place. Sitters, who are the judges of their own readings, (because only they possess all the necessary facts about themselves) can be biased either for or against mediumship and they can be vulnerable to Barnum statements. Mediums can take cues from body language and unconscious signaling or even from the sound of someone’s voice. You can also ruin experiments with faulty scoring.
The sitters need to be properly pre-screened and the mediums need to be capable of doing what they claim they can do. It does no one any good to test using unreliable sitters and incompetent mediums.
But you also need to present the medium with a comfortable and natural atmosphere to do their work and avoid over-designing controls into the process, such as having the sitters all be roughly the same age, from the same area and a similar socioeconomic background. (This adds an unnecessary level of complexity to the reading. Information about age, locations and backgrounds will be harder to distinguish from other sitters.) Other problems include overworking or rushing the medium on any given day.
There is so much that can go wrong.
The First Modern Effort For Mediumship Research
Accounting for all of this can be daunting and requires great attention to detail and a solid experimental design. In 1999, Professor Gary Schwartz of the University of Arizona, conducted the first modern mediumship research experiments, and had partial success(You can read a somewhat critical, but fair peer review of his work here.)
There was a back and forth of criticism and rebuttal with Professor Emeritus Ray Hyman of the University of Oregon, but because of further research and improvements in study design, it’s completely irrelevant at this point. I’ll refer to the Wikipedia stub article, which has the necessary links.
His somewhat flawed mediumship research led to various studies in the early 2000’s, many of which were criticized for the reasons stated above. A skeptic’s study by psychologist Ciaran O’Keeffe and Professor Richard Wiseman for example, put mediums alone in a room for over five hours giving multiple readings, chose sitters that were far too similar in age and background and failed to do trial runs. (Unsurprisingly, they failed.)
Mediumship Experimentation Lessons
Side note: having failed to learn anything from his previous mediumship experiment, and not taking any cues from much better research that had already been published, Professor Richard Wiseman, along with Professor Chris French, together managed to create the most gawd awful mediumship test imaginable. It is better described as a skeptic publicity stunt and is not part of the scientific literature.
The medium had to WRITE ten readings over the course of an eight hour day for a collection of students who were all roughly the same age from the same school. As if that wasn’t bad enough, they blacked out sections of the readings and then handed them over to the students to see if they could pick out whose was whose. The medium, Patricia Putt, who had never claimed to be able to write out readings, unsurprisingly did not get a single match.
A Difficult Learning Period
Other mediumship research studies were criticized for insufficient blinding, pre-screening problems and sensory leakage. The result of all of these false starts was an improvement in the design of mediumship testing. The researchers worked out how to remove the flaws and still provide the mediums with a natural flow. The pre-screening of sitters got better and the researchers learned what sort of blinded testing mediums were capable of.
The Triple Blinded Experiment
By 2006 most of these issues had been solved. Gary Schwartz teamed up with researcher Julie Beischel to create a triple blinded research study. The study shows just how difficult it is to navigate all the different problems associated with testing mediums.
The sitters were pre-screened to ensure that they had actually lost someone they wanted to communicate with and so that they could be trusted to competently complete the trial they were in. They were given both dummy readings and the real ones and asked to rate them. The dummy readings gave the researchers insight into how the person did their scoring so that their responses could be weighted according to the biases they demonstrated, either for or against mediumship and it gave a baseline for chance results.
The mediums were properly screened as well so that the researchers could be sure they were dealing with a competent person. During the test, the medium only dealt with a proxy sitter who was blinded to the actual sitter. When the sitter got their reading results, they were blinded to which one was actually the correct one. The experimenters were blinded to both the medium and the sitter. So the medium, the sitter and the experimenter were all blinded by protocol. Even with all these controls, the mediums still outperformed chance results.
Now that there was finally a proper experiment for others to replicate, various research groups took up the challenge and in 2020 a meta analysis of 14 different studies for a total of 18 different experiments showed that mediums could indeed pass a sufficiently controlled test of their abilities.
The Advancement of Mediumship Research
There was soon a list of certified research mediums to draw upon, which in turn led to other research. Having a group of people in which their psychic ability was confirmed meant that they could be studied for similarities in their physiology, psychology and what mental state they were in while they were giving readings.
What they discovered was that mediums in general experienced good mental health overall, but did worse than average when under emotional distress. This is consistent with highly sensitive people in general. In addition, they definitely entered a different physiological state when they were giving readings. Afterwards they experienced relaxation, clarity and happiness.
There are several things to be learned from this. Criticisms of the initial study designs came from within the parapsychological community, not outside skeptics. Parapsychologists as a group were perfectly capable of seeing the initial flaws in the research and correcting them, just like in any other scientific field. Critics like Ray Hyman were ultimately just noise in the system.
When a couple of skeptics did conduct a study, (Wiseman and O’Keeffe, 2004) their unique experiment was crude and poorly designed and they failed to achieve significant results. They seemed to not have much of a feel for the research and their suggestions for further research were either obvious or unneeded. The progress of mediumship research seems to have not needed their efforts. Wiseman, as far as I’ve been able to discern, has never done a successful psi study of any kind.
One thing I find interesting here is that this relatively small part of parapsychology has flown under the radar, overshadowed by psychologist Dr Daryl Bem’s precognition research. Without the publicity, it hasn’t attracted as much skeptical wrath. It’s a pattern I’ve seen over the years. Skeptics tend to focus the most attention on subjects that reach a wider audience and tend to ignore those that don’t. I will let the reader reach their own conclusions about why that is.
The mediumship studies make an excellent case study for how science works. You try, you fail, you learn, you try again. Once you get it right, it’s time for replications. It’s an evolution of experimental design. And a good study can yield information that informs still more studies, continuing the process of learning through science.