Parapsychology: scientists, methods and evidence
Parapsychology [often referred to as psi (pronounced ‘sigh’)] is a loaded word. For some people it means ghost stories. For others it means “definitely fake.” This section uses the word in a more specific way: as a research tradition that tries to test unusual claims about mind and information using the same basic tools used across science—careful definitions, controlled procedures, and statistics. The point is not to ask you to believe anything. The point is to ask a narrower question: when a claim is tested under good controls, what do the results look like? If the results hold up, that matters. If the results fail under stronger methods, that matters too. Either way, the only honest way forward is to be clear about the experiments. This hub page is a guide to the major ideas, the common experimental approaches, and the scientists who helped build them.
The guide — open any section
What this section is (and isn’t)Think of this site section as a map. It’s a way to find the people, papers, and methods that have shaped modern parapsychology research—especially work that tries to be…
What this section is (and isn’t)
Think of this site section as a map. It’s a way to find the people, papers, and methods that have shaped modern parapsychology research—especially work that tries to be “experiment-first,” not “story-first.” You’ll see researchers who ran laboratory studies, built automated systems to reduce human influence, developed blind judging procedures, and argued over statistics and replication the same way other fields do.
It’s also important to say what this is not. This is not a catalog of paranormal beliefs. It’s not a place where every claim is treated as true. And it’s not a place where criticism is ignored. A field becomes more scientific—not less— when it keeps tightening its methods, publishing disagreements, and making it easier for others to check the work.
If you’re skeptical, you’re the intended audience. The most useful question is never “do you believe it?” The useful question is “how would you test it in a way that rules out ordinary explanations?”
What makes it science?A lot of people assume science is defined by the topic.
What makes it science?
A lot of people assume science is defined by the topic. It isn’t. Science is defined by the rules you follow: you say exactly what you’re measuring, you try hard to avoid fooling yourself, and you let the data be checked.
In parapsychology, the biggest challenge is obvious: if a claim involves information “getting through” in an unusual way, then you have to be extra strict about the ordinary ways information could sneak through. That’s why you’ll see recurring ideas like randomization (so targets can’t be predicted), blinding (so people involved can’t influence outcomes), and predefined statistics (so you don’t decide what “counts” after you already saw the results).
In other words, good parapsychology research doesn’t try to win an argument by telling a convincing story. It tries to earn confidence the boring way: by building procedures where being wrong is possible, and where “normal explanations” have fewer places to hide.
That doesn’t mean the field is settled. It means the field is testable. And “testable” is the line between belief and science.
The main kinds of studiesParapsychology research isn’t one experiment—it’s a family of methods.
The main kinds of studies
Parapsychology research isn’t one experiment—it’s a family of methods. Below are some of the best-known approaches. Each one has its own strengths, its own weaknesses, and its own history of debate.
Ganzfeld and autoganzfeld
The ganzfeld approach is built around a simple idea: reduce normal sensory input and then run a tightly controlled information test. The “reduced input” condition is not the claim by itself—it is a way to simplify the situation so that the task can be defined clearly and scored cleanly.
In many studies, a “sender” looks at a randomly chosen target (an image or video clip), while a “receiver” in the ganzfeld state describes impressions. The critical point is procedural: the target is defined in advance, the timing is fixed, and the outcome is anchored to a concrete matching decision rather than a loose narrative.
Later, a blind judge (or the receiver using a structured rating) tries to match the description to the correct target from a set of possibilities. That matching step is where the “test” lives: it forces the result into a format that can be compared against chance expectations and evaluated consistently across sessions.
Over time, researchers developed more automated versions (“autoganzfeld”) to reduce subtle human influence—like experimenter cues or inconsistent procedures. The purpose of automation here is straightforward: tighten the protocol so that ordinary information pathways have fewer places to enter, and so that the procedure is easier to repeat in the same way.
Dream-ESP and sleep laboratory studies
Dream studies ask whether randomly selected targets can show up in dream reports more often than chance would predict. The central idea is simple: take a dream report as data, define a target set clearly, and then evaluate whether the matching pattern looks different than what would be expected if the target were unrelated.
The challenge is making it a real test instead of a “pattern-finding” exercise after the fact. Without strict structure, it is easy to treat almost any dream as “close enough,” especially when the interpretation is flexible and the target pool is broad.
That’s why good designs rely on clear target pools, careful record keeping, and blind judging so that matching isn’t driven by expectation. Those elements do not guarantee a particular outcome; they make the outcome interpretable by minimizing ambiguity about what was recorded, what counted as the target, and how the match decision was made.
A well-known name in this area is Stanley Krippner, whose work is often discussed in the context of dream laboratories and structured approaches to dream content analysis. In practice, the emphasis is on turning a subjective report into something that can be evaluated under a repeatable method rather than relying on persuasive storytelling.
Mediumship accuracy studies
Mediumship research is often the first thing people think of when they hear “parapsychology,” and it is also one of the easiest areas to do badly. If a medium can see a sitter, get subtle feedback, or learn details indirectly, impressive “hits” can occur without anything unusual happening.
That is why the design problem in this area is not “how dramatic can the reading sound,” but “how well can ordinary information channels and social signaling be blocked.” If those pathways are not controlled, the resulting accuracy claims are hard to interpret because the source of the information is not well constrained.
Modern mediumship research—when it aims to be scientific—leans hard on blinding and information barriers. In a strong design, the medium receives minimal information, the sitter is kept blind to which reading is theirs until rating time, and the data are scored using structured methods that make cueing and cold reading much harder.
The goal of these controls is not to pre-judge the outcome. It is to create a situation where the scoring is anchored and where the result—whether weak, strong, or mixed—can be evaluated without relying on personal impressions about sincerity, charisma, or narrative fit.
Brain, body, and environment
Some researchers approach anomalous experiences from a different angle: instead of asking “did information transfer happen,” they ask “what changes in the brain or body during these experiences?” This shifts the emphasis from a single claim about target-matching to a broader question about measurable correlates that might accompany unusual reports or states.
That leads to studies involving physiology, environmental variables (like geomagnetic conditions), and controlled stimulation. The shared theme is methodological: define what is being measured, collect it in a disciplined way, and evaluate patterns under conditions that are as specified as possible.
This line of work does not automatically answer whether an experience is “paranormal.” Instead, it organizes the question around observable changes—brain, body, and environment—so that competing interpretations can be discussed using recorded measures rather than only personal testimony.
One well-known figure in this general neighborhood is Michael Persinger, whose work is often discussed in relation to neuropsychological and environmental interpretations of unusual experiences. In practical terms, the focus is on whether specific conditions or manipulations are associated with consistent differences in measured responses.
What the evidence looks like in real lifeThe honest answer is: the evidence is complicated.
What the evidence looks like in real life
The honest answer is: the evidence is complicated. Some research lines report small effects that appear across groups of studies, while other analyses argue those effects weaken when you focus only on the most rigorous experiments, or when you correct for issues like publication bias. On top of that, replication in any difficult area is rarely a simple “yes/no.” It depends on the exact protocol, the lab, the analysis plan, and whether the study was truly independent.
This is one reason the best conversations about parapsychology aren’t about single “amazing” studies. They’re about patterns: how results change when controls tighten, how findings look across many experiments, and how researchers respond when critics point out flaws.
Institutions and communitiesAnother common assumption is that parapsychology is just a loose collection of enthusiasts.
Institutions and communities
Another common assumption is that parapsychology is just a loose collection of enthusiasts. Historically, that’s not accurate. Research has been conducted in university-connected labs, independent institutes, and professional communities that maintain conferences and peer discussion. Like many controversial areas, it has also lived under scrutiny—which can be a problem, but can also push methods to improve.
This site focuses less on institutional politics and more on the work itself, but if you’re looking for context—where research groups have existed, what professional organizations do, and how the field has organized over time—see Institutions & Research Communities.
Where to startIf you’re new, the easiest entry point is the people.
Where to start
If you’re new, the easiest entry point is the people. The scientist profiles are written to be readable without a statistics background, while still pointing to sources so you can go deeper.
If you’re skeptical (or just careful), start with method instead of personalities. Methods, controls, and replication standards are the fastest way to understand why results are debated and what “good evidence” would look like in practice.
Whatever your starting point, this section is built around one practical idea: extraordinary claims deserve ordinary clarity. Clear definitions, clear controls, and clear reporting are how you find out what’s real, what’s error, and what’s still unknown.
More from ESP-Nexus: