The Barnum effect: why personality results and horoscopes feel accurate, and how to read any test result

Maybe you have read a personality test result or a daily horoscope and thought, “How is this so me?” The Barnum effect is the tendency to take a vague description that fits almost anyone as an accurate account of you in particular. A 1985 review of the research defined it as the tendency to accept vague, ambiguous and general statements as descriptive of one’s unique personality (Dickson & Kelly, 1985). It is also called the Forer effect, after the psychologist Bertram Forer, who first demonstrated it in an experiment (Poškus, 2014).

Barnum statements come in a few familiar kinds. A 1955 study gave examples: vague ones such as “you enjoy a certain amount of change and variety in life”; double-headed ones such as “you are generally cheerful and optimistic but get depressed at times”; ones true of most people in the group, such as “you find that study is not always easy”; and flattering ones such as “you are forceful and well-liked by others” (Sundberg, 1955, as cited in Dickson & Kelly, 1985). Statements like these can turn up in blood type, zodiac, Chinese zodiac and dream readings, and in the write-ups of personality tests too.

mind.io runs personality tests itself, so this is not someone else’s problem. The write-ups of the mind.io MBTI test and Big Five test can also contain sentences that fit many people to some degree, which can make a result feel more accurate than it is. This guide covers Forer’s 1949 experiment and where the name “Barnum” came from, which descriptions feel most accurate, whether people can tell their real results from fake ones, how belief and culture play in, and how to give any result a second, closer reading.

The first experimental demonstration came from a classroom study the psychologist Bertram Forer published in 1949. As the 1985 review describes it, Forer gave an interest test, the Diagnostic Interest Blank, to 39 students in his introductory psychology class, and a week later handed each of them a personality description. In fact every description was the same, made of vague, ambiguous and general statements that came largely from a newsstand astrology book. The students rated how accurate it was from 0 (poor) to 5 (perfect): the mean was 4.3, nobody rated it below 2, and only 5 students rated it below 4 (Forer, 1949, as cited in Dickson & Kelly, 1985). We could not read Forer’s original paper, so the figures follow that review.

The core of what Forer set out to show with this experiment was that a test cannot be validated by the people tested agreeing that their results fit; he called this the “fallacy of personal validation” (Poškus, 2014). The 1985 review summarises his warning the same way: such personal validation does not distinguish one person’s unique attributes from another’s, so it cannot affirm the validity of the test (Dickson & Kelly, 1985).

The name “Barnum effect” comes from a 1956 paper by the psychologist Paul Meehl. Meehl observed that many psychological test reports bore a disconcerting resemblance to what his colleague Donald Paterson called “personality description after the manner of P. T. Barnum”. He proposed calling “Barnum effect” those pseudo-successful clinical procedures in which descriptions from tests fit the patient largely because they are trivial, and in which any non-trivial but possibly wrong inferences are hidden among statements that are almost bound to be true because they apply to most people; giving such fallacies a name, he wrote, is one of the best ways to make people alert to them (Meehl, 1956). At first the term described a fault on the side of whoever interprets the test; later research, like the 1985 review, uses it for the tendency of the person reading the description to accept it (Dickson & Kelly, 1985). That review introduces P. T. Barnum himself as a famous circus owner whose formula for success was always to have a little something for everybody (Snyder & Shenkel, 1976, as cited in Dickson & Kelly, 1985).

Studies after Forer mostly followed the same steps: people take a personality test, wait while it is scored, receive a description said to come from their test (in most cases the same one for everybody) and rate how accurate it is. In most of these studies, the majority rated the description as a good or excellent account of their own personality (Dickson & Kelly, 1985). In a 1963 study, students rather than a psychologist gave the tests to friends and handed over the same description; 75% still rated it good or excellent, and the mean rating (4.05, 79 people) was no different from an earlier experiment in which a psychologist delivered it (4.38, 57 people). Participants even wrote comments praising the examiner (from the abstract; Ulrich et al., 1963; figures from Dickson & Kelly, 1985).

The Barnum effect is not equally strong for every description. The 1985 review concluded that acceptance depends on how relevant to you the description seems and how favourable it is, and to some extent on the kind of assessment it is said to come from (Dickson & Kelly, 1985).

Being told a description is “for you” strengthened the effect. In a 1972 study, people who were told that the same general description had been derived specifically for them rated it as a more accurate account of their personality than people told it was generally true of people, and later studies repeated the finding (Snyder & Larson, 1972, as cited in Dickson & Kelly, 1985). A result screen that shows your name or says “your personal result” may feel more convincing for the same reason, although these studies compared descriptions labelled “for you” with ones labelled “for people in general” and did not test name-bearing result screens directly.

Flattering descriptions were accepted more readily. Studies that set favourable and unfavourable descriptions against each other in advance found the favourable ones more readily accepted as accurate. Whether the interpreter was a psychologist or a student, or the result came as a computer printout or in handwriting, generally made little difference, except that unfavourable descriptions were more readily accepted from someone of high perceived status (Dickson & Kelly, 1985).

The kind of test showed only a weak effect. The review found a weak trend for descriptions said to come from projective tests to be rated as more accurate than those from other methods, while one study found no statistically significant difference whether the description was said to come from a projective test, handwriting analysis or a horoscope (Dickson & Kelly, 1985).

Looking ahead, the review also raised the possibility that one or two strikingly accurate statements may lead people to accept everything else (Marks & Kammann, 1980, as cited in Dickson & Kelly, 1985). It is presented as a question for research, not an established finding.

Barnum research does not show that people do not know themselves. The 1985 review points out that people accept many Barnum statements because they do fit, and because they have nothing to compare them with. Almost everyone accepts “You have a tendency to be critical of yourself” and “At times you are extraverted, affable, sociable, while at other times you are introverted, wary, reserved” as true of themselves (Bachrach & Pattishall, 1960, as cited in Dickson & Kelly, 1985), but most people would not accept a specific statement such as “You have two eyes, one brown and one grey”. With feedback of a specific nature, the review adds, people often have no trouble telling true from false (Dickson & Kelly, 1985).

How you ask matters too. One study asked separately whether a description was accurate and whether it described the person as unique: students rated the general description as accurate but not as a description of their unique personality, and realised it would fit any of their classmates just as well (Greene, 1977, as cited in Dickson & Kelly, 1985). That seems at odds with other studies in which people rated a description as truer of themselves than of people in general, and the review suggests the two approaches may have measured different things. It concluded that people can tell accuracy from uniqueness when asked directly, but that it is unclear whether they do so when not asked; the fact that people often treat such accuracy ratings as confirmation of the test, it added, suggests they do not (Dickson & Kelly, 1985).

Some studies mixed in real scores. In one, 75 people (56 military aviation cadets and 19 graduate students) who had taken the Norwegian translation of the NEO PI-R, a Big Five personality inventory, received a randomly generated score profile instead of their own and rated their agreement on a 7-point scale. The closer the random scores happened to be to their real ones, the more they agreed (correlations from .08 to .64 depending on the trait), and they identified and rejected random scores far from their real ones; the authors saw this as working against the Barnum effect (from the abstract; Andersen & Nordvik, 2002).

In a 2008 study, 52 college students were shown one real and one bogus result and asked to pick their own. They picked their real Big Five (NEO-FFI) profile at a greater-than-chance rate but could not pick their real summary from a computer-generated astrological natal chart; yet when they rated accuracy, a Barnum effect showed up for both (from the abstract; Wyman & Vyse, 2008). One way to read this is that being able to pick out your own result and feeling that a result fits can come apart.

A 2014 study that randomly assigned 359 Lithuanian university students to four groups, and analysed the 243 who came back for the second session, gives a sharper contrast. Two to three weeks after taking a Big Five inventory (NEO PI-R), each person received a five-sentence personality summary: one that matched their real scores, one as opposite to their real profile as possible, one that painted every trait favourably, or one that called every trait average. On a 7-point accuracy scale, the real, favourable and average summaries all had a median of 6 with no difference between them; only the opposite summary was rated lower, with a median of 5 (Poškus, 2014). Real results, in other words, were not rated as more accurate than flattering ones or average ones that could be said to anyone. Still, in the favourable-summary group, ratings of individual sentences were mostly related to people’s actual trait scores (openness was the exception), and the author read the ratings as containing some objective judgement, not only the Barnum effect (Poškus, 2014).

A 1989 study comparing believers in astrology with sceptics found that believers were more likely overall to accept descriptions as accurate. Favourable descriptions were generally rated as more accurate, with one exception: believers accepted a negative astrological description as much as a positive one. Sceptics as well as believers rated a description as more accurate when it was said to come from astrology than when the identical description was not, and sceptics who received a favourable horoscope, unlike every other group, became significantly more positive about astrology (from the abstract; Glick et al., 1989).

Differences by culture were not clear. A 2009 study gave 130 Westerners and 128 Chinese nationals a Barnum profile supposedly derived from either Western or Chinese astrology. A first analysis seemed to support a universal Barnum effect, but it disappeared once gender, age, general education and psychological knowledge were controlled for. There was little support for cultural differences in astrological belief or in susceptibility to the Barnum effect, and whether the profile was said to come from Western or Chinese astrology made no difference, although Chinese nationals who believed in astrology rated their own profile as more accurate for people in general than Chinese sceptics did (from the abstract; Rogers & Soule, 2009).

People can also move towards the result they are given. In a 1995 Japanese experiment, 64 students at a women’s university took either a pop psychology quiz taken from popular books or an academic personality test, and were then told at random, falsely, that they were extraverted or introverted. Those told they were extraverted rated themselves as more extraverted than those told they were introverted. In a waiting room they more often started talking to a stranger (a confederate), and both the confederate and raters who watched the video saw them as more extraverted. The effect on behaviour did not differ between the pop quiz and the academic test, and agreement with the result was above the midpoint of 4 on the 7-point scale overall, with no significant difference between conditions (Sakamoto et al., 1995).

The authors noted that all participants were female university students and that more checks are needed to know whether brief behaviour in a lab carries over into daily life. They also acknowledged that the “extraverted” feedback was rated as more desirable than the “introverted” feedback, so an explanation through desirability and mood remains possible, although they themselves thought it less likely (Sakamoto et al., 1995). A 1974 study described in the 1985 review likewise found that people given an astrological profile describing them as high in dominance and change later rated themselves higher on those qualities in a personality inventory than people given a low profile, though whether their behaviour changed was not clear (Delaney & Woodyard, 1974, as cited in Dickson & Kelly, 1985).

Psychology students are not immune either. Researchers at a French university had 2,269 undergraduate psychology students, across five cohorts, take a fake online test before their first psychometrics course: 85% rated its effectiveness above the neutral point, and 77% said they would be willing to use it as future psychologists (from the abstract; Gonthier & Thomassin, 2025). According to the 1985 review, one study found senior psychology students rated Barnum descriptions less favourably than younger students (Greene, 1977, as cited in Dickson & Kelly, 1985), while most studies found no significant link between acceptance and sex. The review also noted that most studies so far had been done with university students (Dickson & Kelly, 1985).

The central lesson of Barnum research is that the feeling that a result fits does not tell you whether the test is accurate. As Forer warned, such personal validation does not distinguish one person from another (Dickson & Kelly, 1985), and a 1979 study concluded that accepting personality feedback is not evidence that a test is valid, nor is rejecting it evidence that the test is invalid (from the abstract; Layne & Michels, 1979). The 1985 review adds that genuine or bogus results that contain Barnum statements will feel accurate, giving an illusion of validity, and that people who find a result accurate gain faith in the test (Dickson & Kelly, 1985).

It works the other way too: a result that feels right is not proof that the test is bogus. In the studies above, people picked their real Big Five result better than chance (Wyman & Vyse, 2008) and rejected scores far from their real ones (Andersen & Nordvik, 2002). Whether a test can be trusted is judged not by feel but by studies that check whether the same person gets similar results again (reliability) and whether the test measures what it sets out to measure and relates meaningfully to other evidence (validity). How the MBTI fares on these standards is covered in MBTI accuracy, and how well supported the popular personality systems are in MBTI, blood types, zodiac signs and saju.

Readings based on blood type, birth date, looks, palm lines, handwriting or dreams, such as blood type personality, zodiac signs, Chinese zodiac animals, face reading, palm reading, graphology and dream interpretation, cannot be judged by the feeling that they fit either. Whether they really relate to personality or the future is a question for research. The studies are set out in blood type and personality, zodiac signs and personality, Chinese zodiac and personality, face reading, palm reading, handwriting and dream interpretation, and how to read saju differently from personality tests in saju vs personality tests. Why tarot card readings can feel accurate, and the few studies that tested them, are covered in the tarot guide. Typing people as left-brained or right-brained borrows the language of brain science, but it too has to be judged by research, not by the feeling that it fits. The two halves of the brain do divide some of the work, yet the studies set out in left-brained or right-brained give no grounds for sorting people into those two types. mind.io’s blood type, zodiac and saju pages present these readings as popular culture to enjoy.

The same goes for mind.io results. When an MBTI type description or a Big Five interpretation makes you nod along, part of that may be the Barnum effect. So it helps to look at the numbers calculated from your answers as well as the descriptions: the mind.io MBTI test shows how far, in percent, you lean on each of the four axes and which axes are close to 50:50, and the Big Five test shows a score for each of the five traits.

The ideas below are suggestions drawn from Barnum research; the methods themselves have not been tested. Treat them as a checklist for reading results.

First, ask whether it would fit a friend. Read the result one sentence at a time and ask, “Would this be true of a friend or someone in my family too?” A sentence that fits most people tells you little that is new about you. The idea comes from the finding that people separated accuracy from uniqueness when asked about both (Greene, 1977, as cited in Dickson & Kelly, 1985).

Second, read the descriptions of other results. If many sentences in MBTI types that are not yours, or in blood type profiles other than your own, also make you nod, those sentences do not set you apart from anyone. Types one letter apart are compared side by side in look-alike MBTI types.

Third, look for sentences that could be wrong. A double-headed sentence such as “at times extraverted, at times introverted” is hard to get wrong. A result tells you something when a sentence that might not have fitted, such as “in a group, you tend to lead the conversation”, turns out to fit. As the eye-colour example in the 1985 review suggests, the more specific a statement, the easier it is to tell whether it is true (Dickson & Kelly, 1985).

Fourth, look at scores and close calls rather than the description. For an MBTI result, check which axes were close to 50:50; for the Big Five, which traits sit near the middle. A result near a boundary can change on a retest. Why results change and how to read them is covered in detail in MBTI accuracy.

Fifth, retest after a while, or compare with a different test. Whether you get the same direction a few weeks later on an ordinary day, neither exhausted nor elated, and whether tests that work differently, such as the MBTI test and the Big Five test, point the same way, gives you more to go on than a single result. How to read the two together is in MBTI vs Big Five. Asking someone who knows you well which description sounds more like you may help too.

Sixth, enjoy the pages that are for fun as fun. Hunting for the lines that fit in a blood type or zodiac profile and talking it over can be enjoyable. Just don’t use such results to judge people or to make important decisions about dating or hiring. It also helps to remember that the more flattering a result, the more readily it tends to be believed (Dickson & Kelly, 1985).

Q. Are the Barnum effect and the Forer effect the same thing? A. Yes, they name the same phenomenon. Forer demonstrated it in a 1949 experiment and called it the “fallacy of personal validation” (Poškus, 2014); the name “Barnum effect” was proposed by Paul Meehl in 1956 (Meehl, 1956), who first used it to criticise trivial descriptions in test reports.

Q. What are some examples of the Barnum effect? A. Typical examples are statements that fit almost anyone to some degree, such as “you are generally cheerful and optimistic but get depressed at times”, “you find that study is not always easy” and “you have a tendency to be critical of yourself” (Dickson & Kelly, 1985). A horoscope’s “a change is coming” or a blood type profile’s “usually calm, but sometimes sensitive” are the same kind.

Q. Is an accurate-feeling MBTI result just the Barnum effect? A. Partly, it may be. Type descriptions can contain sentences that fit many people, which can make the sense of fit bigger than it really is. But the MBTI calculates scores from the items you answer, and how consistent the official MBTI’s scores are and what they relate to has been studied separately. The mind.io MBTI test is not the official instrument but MBTI-style items written for the same four preference axes, so that research does not validate the mind.io test itself; see MBTI accuracy.

Q. Is the Barnum effect why blood type or zodiac personality feels so accurate? A. It is one of the reasons given. Believers in astrology accepted descriptions as more accurate, and even sceptics rated an identical description as more accurate when it was said to come from astrology (Glick et al., 1989). The research on each is in blood type and personality and zodiac signs and personality.

Q. If a result feels accurate, is the test accurate? A. That alone does not tell you. One study concluded that accepting personality feedback is not evidence that a test is valid (Layne & Michels, 1979). Nor does a sense of fit mean the test is wrong. Look for studies of whether people get similar results again (reliability) and whether the test measures what it claims to (validity).

Q. Are some people more prone to the Barnum effect? A. Most studies found no significant link between acceptance and sex. There is some evidence of links with personality characteristics such as an external locus of control, authoritarianism and the need for approval (Dickson & Kelly, 1985), but it can happen to anyone: even among psychology undergraduates, 85% rated a fake test’s effectiveness above the neutral point (Gonthier & Thomassin, 2025).

Q. How can I tell whether a result is fake? A. Check whether the sentences would fit a friend, whether they contain specific claims that could be wrong, and whether a retest or a different test agrees. These are suggestions drawn from Barnum research rather than methods tested in studies.

References: Dickson, D. H., & Kelly, I. W. (1985), The “Barnum effect” in personality assessment: a review of the literature, Psychological Reports 57(2), 367-382, https://doi.org/10.2466/pr0.1985.57.2.367; Forer, B. R. (1949), The fallacy of personal validation: a classroom demonstration of gullibility, Journal of Abnormal and Social Psychology 44(1), 118-123, https://doi.org/10.1037/h0059240 (cited in Dickson & Kelly, 1985); Sundberg, N. D. (1955), The acceptability of “fake” versus “bona fide” personality test interpretations, Journal of Abnormal and Social Psychology 50(1), 145-147, https://doi.org/10.1037/h0044374 (cited in Dickson & Kelly, 1985); Poškus, M. S. (2014), A new way of looking at the Barnum effect and its links to personality traits in groups receiving different types of personality feedback, Psichologija 50, 95-105, https://doi.org/10.15388/psichol.2014.50.4893; Meehl, P. E. (1956), Wanted—a good cookbook, American Psychologist 11(6), 263-272, https://doi.org/10.1037/h0044164 (read in its reprint, Problems in the actuarial characterization of a person, Minnesota Studies in the Philosophy of Science 1, 205-222, https://hdl.handle.net/11299/184265); Snyder, C. R., & Shenkel, R. J. (1976), Effects of “favorability,” modality, and relevance on acceptance of general personality interpretations prior to and after receiving diagnostic feedback, Journal of Consulting and Clinical Psychology 44(1), 34-41, https://doi.org/10.1037/0022-006X.44.1.34 (cited in Dickson & Kelly, 1985); Ulrich, R. E., Stachnik, T. J., & Stainton, N. R. (1963), Student acceptance of generalized personality interpretations, Psychological Reports 13(3), 831-834, https://doi.org/10.2466/pr0.1963.13.3.831; Snyder, C. R., & Larson, G. R. (1972), A further look at student acceptance of general personality interpretations, Journal of Consulting and Clinical Psychology 38(3), 384-388, https://doi.org/10.1037/h0032899 (cited in Dickson & Kelly, 1985); Marks, D., & Kammann, R. (1980), The psychology of the psychic, Prometheus Books (cited in Dickson & Kelly, 1985); Bachrach, A. J., & Pattishall, E. G. (1960), An experiment in universal and personal validation, Psychiatry 23(3), 267-270, https://doi.org/10.1080/00332747.1960.11023225 (cited in Dickson & Kelly, 1985); Greene, R. L. (1977), Student acceptance of generalized personality interpretations: a reexamination, Journal of Consulting and Clinical Psychology 45(5), 965-966, https://doi.org/10.1037/0022-006X.45.5.965 (cited in Dickson & Kelly, 1985); Andersen, P., & Nordvik, H. (2002), Possible Barnum effect in the five factor model: do respondents accept random NEO Personality Inventory-Revised scores as their actual trait profile?, Psychological Reports 90(2), 539-545, https://doi.org/10.2466/pr0.2002.90.2.539; Wyman, A. J., & Vyse, S. (2008), Science versus the stars: a double-blind test of the validity of the NEO Five-Factor Inventory and computer-generated astrological natal charts, Journal of General Psychology 135(3), 287-300, https://doi.org/10.3200/GENP.135.3.287-300; Glick, P., Gottesman, D., & Jolton, J. (1989), The fault is not in the stars: susceptibility of skeptics and believers in astrology to the Barnum effect, Personality and Social Psychology Bulletin 15(4), 572-583, https://doi.org/10.1177/0146167289154010; Rogers, P., & Soule, J. (2009), Cross-cultural differences in the acceptance of Barnum profiles supposedly derived from Western versus Chinese astrology, Journal of Cross-Cultural Psychology 40(3), 381-399, https://doi.org/10.1177/0022022109332843; Sakamoto, A., Miura, S., Sakamoto, K., & Mori, T. (1995), A self-fulfillment phenomenon due to outcome feedback of pop psychological tests: an experiment and a survey of female undergraduate students [in Japanese], Japanese Journal of Experimental Social Psychology 35(1), 87-101, https://doi.org/10.2130/jjesp.35.87; Delaney, J. G., & Woodyard, H. D. (1974), Effects of reading an astrological description on responding to a personality inventory, Psychological Reports 34, 1214 (cited in Dickson & Kelly, 1985); Gonthier, C., & Thomassin, N. (2025), Getting students interested in psychological measurement by experiencing the Barnum effect, Teaching of Psychology 52(2), 183-192 (online 2024), https://doi.org/10.1177/00986283241240454; Layne, C., & Michels, P. J. (1979), Inventory responding as a model of people’s acceptance of personality interpretations, Journal of Personality Assessment 43(5), 509-513, https://doi.org/10.1207/s15327752jpa4305_13