Are Mediums Real? What Happened When a Scientist Tested Them in a Lab?
I want to introduce you to a piece of research that sits especially close to home for me, because it takes the very thing I do for a living and asks whether it can hold up under the same scientific scrutiny we would apply to any other claim. Dr. Gary Schwartz is a psychologist who spent decades at Yale and Harvard before joining the University of Arizona, holding professorships across psychology, medicine, neurology, psychiatry, and surgery. In the early 2000s, he turned that formidable academic credibility toward a question most of his peers considered far too risky to touch: can mediumship actually be tested, under controlled conditions, the way any other scientific claim is tested?
I have spent this entire series researching mystics, trance mediums, near-death researchers, and reincarnation scientists, and Schwartz's work occupies a unique place in that lineup. He is not describing his own spiritual experience, the way Swedenborg or Anita Moorjani did. He is not interviewing survivors of a medical crisis, the way Moody and van Lommel did. He built laboratory experiments specifically designed to test whether mediums like the ones I work alongside in this field can actually perceive information about the deceased that they have no ordinary way of knowing.
I think it is worth pausing on just how much professional risk this represented for Schwartz personally. He came to this research already established, with a doctorate from Harvard and faculty positions at Yale behind him, a career built on conventional psychophysiology research that had nothing to do with mediumship or the paranormal. Turning his academic attention toward testing psychic claims put all of that credibility on the line. As he has acknowledged in his own writing, he approached his early findings with real caution and even discomfort, aware that summarizing and standing behind his results honestly might mean concluding that skeptics had been wrong about something he himself had once assumed was settled. That is not a comfortable position for any scientist to put themselves in voluntarily, and I think it says something about how seriously he took the question that he did it anyway.
How the research began
Schwartz's interest in mediumship research grew out of his broader work studying consciousness and energy at the University of Arizona, where he directs the Laboratory for Advances in Consciousness and Health. Rather than starting from a position of belief, he approached the question as a testable hypothesis. If mediums could genuinely perceive information from a deceased person, that ability should show up as a measurable, statistically significant pattern under conditions designed to rule out cold reading, guessing, and information a medium might pick up from a sitter's body language or verbal cues.
He assembled a group of mediums with established public reputations, including figures like John Edward, George Anderson, and Suzane Northrop, and began running structured experiments comparing what these mediums reported against what a control group of ordinary people, with no claimed mediumship ability, could produce by simple guessing. His early studies, published in his 2002 book The Afterlife Experiments, reported striking results. In one often-cited experiment, a control group of 68 University of Arizona students, given the same basic setup and simply asked to guess, produced an average accuracy of 36 percent. The mediums in the study ranged from 77 to 95 percent accuracy, averaging 83 percent. Schwartz calculated that a gap this large between trained mediums and control guessing was statistically unlikely to occur by chance alone, describing the odds as less than one in ten million.
The design of these early experiments deserves a closer look, because it shaped everything that came after it, including the criticism. Sitters, the people seeking information about a deceased loved one, were present in the room but instructed not to speak, respond verbally, or offer any confirming or denying reaction during the reading itself, so a medium could not simply watch for a flicker of recognition and steer the reading toward it. Mediums were given only a first name to work from, nothing else, and asked to report whatever impressions came to them about the deceased person's appearance, personality, cause of death, and any messages they felt were being communicated. Afterward, sitters rated the accuracy of specific statements on a structured scale, which is how Schwartz arrived at his percentage comparisons between mediums and the control group.
Tightening the controls: the VERITAS and SOPHIA programs
Schwartz was aware from the start that his early experiments would draw serious criticism, and they did, particularly around how well sitters, mediums, and researchers were kept separated from one another during testing. In response, he and his colleague Dr. Julie Beischel developed increasingly rigorous protocols designed to close those gaps entirely. Their work evolved into what became known as the VERITAS Research Program and later the SOPHIA Research Program, both aimed at testing mediumship under conditions that made cheating, unconscious cueing, or cold reading as close to impossible as a laboratory setting can achieve.
The core innovation was what researchers call a triple-blind protocol. The medium has no contact with the sitter and knows nothing about them beyond perhaps a first name. The sitter is not present during the reading at all, removing any possibility of picking up cues from body language, reactions, or tone of voice. And the researcher scoring the results does not know which reading belongs to which sitter until after the scoring is complete. Under this design, a medium is asked detailed questions about a deceased person's appearance, personality, activities, and cause of death, with no sitter in the room to react to, confirm, or unconsciously guide the reading in any direction.
Where the research stands today
When Schwartz's academic position at the University of Arizona ended, Dr. Julie Beischel continued this line of research independently, founding the Windbridge Institute for Applied Research in Human Potential, later renamed Windbridge Research. Windbridge developed an eight-step certification process for research mediums and continued refining the blinded protocols Schwartz and Beischel had pioneered together, building what is now one of the most methodologically careful ongoing research programs into mediumship anywhere in the world.
I find it meaningful that this research did not stop once Schwartz's academic funding shifted. Beischel has continued this work for years, certifying research mediums and publishing findings aimed specifically at bereaved families and, notably, at medical and psychological professionals who work with grief, treating mediumship reading accuracy as a genuinely testable scientific question rather than something to simply accept or dismiss on faith.
Windbridge's certification process is itself worth describing, because it illustrates how far the field moved from Schwartz's original, looser experimental design. A candidate research medium must pass through eight distinct stages, including screening interviews, blinded practice readings scored against known information, and ongoing accuracy testing, before being certified to participate in formal research studies. This is a meaningfully higher bar than simply claiming mediumship ability publicly. It represents an attempt to build a genuine research infrastructure around this question, one with standards, replication, and ongoing quality control, the same basic scientific scaffolding any other field of study would require before taking its claims seriously.
Why the skepticism deserves a fair hearing
I want to be honest about something, because I think honesty is what makes this kind of research worth taking seriously rather than dismissing as advocacy. Schwartz's work has drawn substantial criticism from within the scientific community, and some of it is worth engaging with directly rather than glossing over.
The psychologist and magician Ray Hyman published a detailed critique arguing that Schwartz's early experiments, particularly the widely cited HBO experiment, had significant control weaknesses, including the possibility that a name alone, given to a medium ahead of a reading, carries more identifying information than it might seem to on the surface. Other critics have pointed out that the sample sizes in several of Schwartz's published studies were relatively small, and that with enough experiments conducted across a field, some will produce results that look statistically striking purely by chance, an issue researchers call the multiple comparisons problem. These are legitimate scientific concerns, and I think a medium's credibility, including my own view of this research, is actually strengthened by acknowledging them honestly rather than treating Schwartz's early results as beyond question.
What I think is fair to say is this: the criticism pushed the field to improve, not abandon, its methods. The move from Schwartz's original experiments toward the triple-blind VERITAS and SOPHIA protocols, and now Windbridge's ongoing certification and testing process, represents exactly the kind of self-correction good science is supposed to produce. The core claim, that trained mediums can produce statistically significant accuracy under blinded conditions well above what chance guessing predicts, has been tested repeatedly with increasingly tight controls, and has continued to hold up, even as the specific numbers and interpretations remain genuinely debated.
I want to be clear that this remains a contested area of science, and I do not think it does the field any favors to pretend otherwise. Mainstream psychology and neuroscience largely remain unconvinced, and the criticisms raised by researchers like Ray Hyman have never been fully resolved to the satisfaction of skeptics. What I find genuinely valuable, regardless of where the broader scientific consensus eventually lands, is that this question has been treated as answerable at all, through actual experimental design rather than simply asserted or dismissed based on prior assumption. That alone sets Schwartz's contribution apart from most conversations about mediumship, which tend to stay at the level of belief versus disbelief rather than evidence versus counter-evidence.
What this means for how I think about my own work
I do not do readings in a laboratory, and I do not claim scientific certification the way a Windbridge-certified medium does. But I find real value in knowing this research exists, because it gives language and structure to a question every working medium eventually has to sit with honestly: what actually separates a real, evidential detail from something vague enough that it could apply to almost anyone?
Schwartz's research, and the work that grew out of it, suggests this distinction is not just a matter of faith or reputation. It is something that can be studied, measured, and tested, imperfectly, with real limitations, but tested nonetheless. That matters to me, both as someone who wants to hold my own work to a genuine standard, and as someone who wants the people I read for to trust that this field takes its own credibility seriously enough to withstand real scrutiny.
If you have ever wondered whether mediumship is something science has simply ignored, I hope this shows you otherwise. It has been studied, directly, by researchers with serious academic credentials willing to risk their own reputations to ask the question properly. The answers remain contested, as they should in any genuinely open scientific inquiry. But the question itself has been asked, carefully, again and again, for more than two decades now.
I think about what this means for the person sitting across from a medium for the first time, uncertain and a little guarded, wondering whether any of it is real. Schwartz's research does not settle that question for them in advance. No study could. What it offers instead is context, the knowledge that the underlying claim has been taken seriously enough by credentialed scientists to build careers, laboratories, and certification programs around testing it properly. That does not replace the experience of sitting with a medium yourself and deciding what you make of it. But it does mean you are not the first person to ask whether any of this can be trusted, and you are certainly not the only one who has demanded real evidence before believing it.
Ready to experience a reading?
If you're feeling drawn to connect with a loved one in spirit, I'd be honored to walk that path with you. You can book a mediumship reading at aperfectsoul.com.