Why is this question hard to answer outright?
The question looks simple: do people who believe in the paranormal rate a divination session as more accurate than non-believers? Answering it requires separating two different things. One is how the sitter feels about the session. The other is whether the session actually said anything true. Most existing research measures only the first, and even there the results are not consistent.
At least two experimental studies bear directly on this, and they disagree. Both are small. Neither is large enough to settle anything, and there is no study of online tarot in Vietnam.
The 2001 study: believers rated the reading as more accurate, even with reversed cards
Matthew Hughes and colleagues published a short experiment in The Journal of Social Psychology in 2001 using a reversed-card design. Participants were shown a card reading, but one group received a reading matching the drawn cards while another received a reading with the meanings inverted. They then rated the accuracy of the reading.
The result: people with paranormal belief rated the reading as more accurate, regardless of whether they received the true or the inverted version. In other words, whether the reading matched the cards did not seem to shift this group's judgment. What shifted their judgment was how strongly they believed in the paranormal beforehand.
One caveat: this is level B evidence, meaning a small study, not widely replicated, and still contested. Our archive holds only a catalogue card for this study, not the full text, so we do not cite the sample size or specific figures.
The 2025 study: 67 people, results nearly the opposite
Thomas R. Brooks, Magdalena Krow and Daniel Chadborn published a quantitative study in 2025 in the Journal of Articles in Support of the Null Hypothesis on tarot, paranormal belief and fantasy proneness. They hypothesised that people high in paranormal belief and high in fantasy proneness would rate a tarot reading as more accurate and more useful.
The sample was 67 people (mean age 21.48; standard deviation 5.12), recruited at a campus festival at a university in the southwestern United States over two consecutive years, from autumn 2022 to autumn 2024. Most were women (62.7%), followed by men (25.4%) and non-binary people (11.9%). Two experienced readers read for each participant, who then answered a short questionnaire covering demographics, the reading experience, and the psychological variables of interest.
Perception of the reading was measured with two separate questions: rated accuracy and rated usefulness, both on a 7-point Likert scale. Paranormal belief was measured with Tobacyk's Revised Paranormal Belief Scale, and fantasy proneness with Merckelbach and colleagues' 25-item Creative Experiences Questionnaire.
The result: no statistically significant relationship between paranormal belief or fantasy proneness and rating the reading as accurate (F[8, 66] = .78, p = .62, R2 = .10) or useful (F[8, 66] = .63, p = .75, R2 = .08). The team retained the null hypothesis, meaning no relationship.
Kind of knowledge: DocumentedHistorical record. Backed by a document, a date and a source; the note number beside it points to that source. The five labelsThere was one small exception: a weak relationship between rating the reading as useful and belief in precognition1. This is slightly odd, since belief in precognition should attach more to perceived accuracy than to usefulness, and the team acknowledged having no clear explanation.
Why could both results be right?
The answer lies in the fact that the two studies measured different things. Hughes and colleagues measured how accurate participants judged the reading to be, and manipulated the cards to see whether inverted meanings changed the judgment. Brooks and colleagues measured the felt perception of a real reading, with real readers, and looked for a link between that perception and paranormal belief measured beforehand.
One important difference: Hughes and colleagues found paranormal belief predicted accuracy ratings, while Brooks and colleagues found no such link. Both could be right, because the contexts differ. In the Hughes experiment, participants knew they were taking part in a study about fortune telling, which may have primed a response script. In the Brooks study, participants received a reading at a festival, a more natural setting, and may have come to the reading out of curiosity rather than firm belief.
The difference could also lie in measurement. Hughes measured the perceived accuracy of the reading in general. Brooks measured two separate things, accuracy and usefulness, and found a weak link only with usefulness. Separating these two concepts is an important contribution: a reading can be seen as useful without necessarily being seen as accurate, or the reverse.
Why are 'accurate' and 'useful' two different questions?
Someone can leave a tarot reading and say: 'This helped me look again at what is going on in my life.' That is a usefulness judgment. The same person may not say: 'The cards know what will happen.' That is an accuracy judgment in the predictive sense. The two statements do not exclude each other, but neither implies the other.
In the Brooks study, neither statement was tied to paranormal belief, except for a weak link between usefulness and belief in precognition. This suggests that for many people tarot may have become a tool for self-reflection rather than a tool of prediction. The team connected this to the commercialisation of spirituality through the New Age movement, which has pulled tarot gradually away from its occult origins.
Kind of knowledge: UnprovenNot enough evidence. Someone has claimed it, but no source or study is strong enough to confirm it. The five labelsThe history of tarot supports this reading. Tarot began in mid-15th-century Italy as a card game, then followed European migrants to America and gradually became a divination tool in Western occultism. Decks such as Tarot de Marseille and Rider-Waite are used for cartomancy. In other words, using tarot for divination is a later layer of meaning, not its original nature.
Why does the feeling of 'rightness' not prove the cards know anything?
There is a well-studied psychological mechanism that helps explain why a reading can feel right without the cards knowing anything. Ray Hyman, in the chapter 'Cold Reading: How to Convince Strangers That You Know All About Them' in the book Pseudoscience and Deception edited by B. Farha in 2013, describes the cold reading technique: the reader uses general statements, watches reactions, and adjusts gradually to create the impression of knowing a stranger well.
This is level A evidence, meaning consensus among researchers. Hyman is a respected psychologist in this field, and cold reading has been described and tested across many studies. We have a separate article on this technique: Cold reading: the technique that makes a vague pronouncement sound like it hits your private life.
A related mechanism is the Barnum effect, also called the Forer effect. Peter Glick and colleagues published a study in Personality and Social Psychology Bulletin in 1989 showing that believers in astrology accepted astrologically labelled descriptions more readily. This is level A evidence, meaning consensus among researchers. The result suggests that with the same general description, astrology believers tend to see it as more true of themselves.
One distinction matters: this is an effect of belief on judgment, not evidence that astrology or tarot predicts correctly. Belief changes how people read and remember information, and so changes their judgment. That effect is real, and it says nothing about whether the cards know anything.
Where the evidence is still missing
First, no study is large enough. The Brooks study had 67 people, and the authors themselves acknowledged the narrow scope, intended only to open the way for later research. The Hughes study is also small, and our archive holds only a catalogue card for it.
Second, there is no study of online tarot in Vietnam. Both studies were conducted abroad, on non-Vietnamese groups. No conclusion can be drawn for Vietnamese people without research on Vietnamese people.
Third, both studies measured the sitter's perception, not the actual accuracy of the reading. There is no way to measure the actual accuracy of a prediction about the future, because the future has not happened. This is an inherent limit of all research of this kind.
Fourth, studies differ in how they measure paranormal belief. Hughes used one scale, Brooks used Tobacyk's Revised Paranormal Belief Scale. Different scales can produce different results.
Comparison of the two studies
| Feature | Hughes et al. (2001) | Brooks et al. (2025) |
|---|---|---|
| Journal | The Journal of Social Psychology | Journal of Articles in Support of the Null Hypothesis |
| Sample size | Not in the catalogue card | 67 people |
| Mean age | Not in the catalogue card | 21.48 (SD = 5.12) |
| Gender | Not in the catalogue card | Women 62.7%, men 25.4%, non-binary 11.9% |
| Location | Not in the catalogue card | A university in the southwestern United States |
| Data collection period | Not in the catalogue card | Autumn 2022 to autumn 2024 |
| Design | Reversed card meanings | Survey after a real reading |
| Main result | Believers rated the reading as more accurate, regardless of the true or inverted version | No relationship between paranormal belief and perceived accuracy or usefulness |
| Evidence level | B (not replicated) | B (small sample) |
So what is the answer to the main question?
Kind of knowledge: UnprovenNot enough evidence. Someone has claimed it, but no source or study is strong enough to confirm it. The five labelsThe evidence is thin and contradictory. A small 2001 study found paranormal believers rated a reading as more accurate, even when the cards were reversed. A small 2025 study found the felt perception of a tarot session barely related to paranormal belief. Both are level B, meaning not widely replicated and still contested.
What holds up better is that feeling 'right' does not prove the cards know anything. Hyman's cold reading technique and the Barnum effect studied by Glick and colleagues show that this feeling can be produced by how questions are framed, how reactions are observed, and the listener's own tendencies. These are real effects, arising from human psychology, not from any predictive power in the cards.
But one thing should be said clearly: the fact that this feeling can be explained psychologically does not mean it is worthless to the person experiencing it. If a reading helps someone look again at their situation, ask themselves questions, or find a direction, that value is real to them. It is usefulness value, not predictive value. The two do not exclude each other, but neither implies the other.
Frequently asked questions
Do paranormal believers rate a divination session as more accurate than non-believers?
Research results are not consistent. Hughes and colleagues (2001) found they do, Brooks and colleagues (2025) found they do not. Both are small studies, not widely replicated. No firm conclusion can be drawn from these two studies.
Why do the two studies reach opposite results?
Possibly because they measured different things. Hughes measured rated accuracy of the reading, and manipulated the cards to see whether inverted meanings changed the judgment. Brooks measured the felt perception of a real reading, with real readers, and looked for a link with paranormal belief measured beforehand. Context, measurement, and scales all differ.
Does a tarot reading feeling very accurate mean the cards know something?
No. The feeling of accuracy can be produced by cold reading, the Barnum effect, and the listener's own tendencies. Ray Hyman described how cold reading creates the impression of knowing a stranger well, and Peter Glick and colleagues found astrology believers accepted astrologically labelled descriptions more readily. These are real psychological effects, and they do not prove the cards have predictive power.
