The Barnum Effect
A personality description written to fit almost anyone is routinely read as a precise account of one particular person. The finding takes its common name from P. T. Barnum, on the strength of a maxim about a good act having something for everybody that is only doubtfully his, and its technical name from Bertram R. Forer, the psychologist who gave it a controlled classroom demonstration in 1949. Both names now attach to the same result: a sketch built from broad, mostly favourable, near-universally applicable statements is accepted as an individualised account of a specific person’s character, including by the person it was never really written about.
Forer’s classroom demonstration
Forer gave thirty-nine students in his own introductory psychology class a questionnaire he called a Diagnostic Interest Blank, told them it would produce an individualised personality sketch, and a week later handed each of them a typed page bearing that student’s own name. Every one of the thirty-nine pages carried the same thirteen statements, among them a claim about needing other people’s liking and admiration and a claim about security as a major life goal 1. Asked to rate, on a scale of zero to five, how well the sketch described their own personality, students rated it highly; on average they marked just over ten of the thirteen statements true of themselves.
Correction
Commonly saidForer's students rated the sketch as an accurate description of themselves at a mean of 4.3 out of 5.
The figure is close, and the story around it is right; what has slipped is which of two ratings it belongs to. Forer collected two separate ratings that week, both on a zero-to-five scale, and they answer different questions: one asks how well the sketch describes the student’s own personality, the other asks how good the diagnostic instrument itself is at revealing personality. Computed directly from Forer’s own published frequency table, the mean rating of the sketch is 4.26 and the mean rating of the instrument is 4.31 1. The number quoted everywhere is a rounding of one or the other, the two run together as though they had answered the same question.
There is a small irony: the single most-cited demonstration that people repeat a plausible-sounding statement without checking it is itself repeated, more than seventy years on, with an unchecked number.
One further detail belongs to the paper itself. Forer’s own footnote to the thirteen statements records where they came from: they “came largely from a newsstand astrology book,” and he adds that he had not, at the time, seen a similar sketch already circulating on the lecture circuit 1. The demonstration that gave the effect its name was assembled out of astrological material, by the psychologist who first isolated the problem, and the fact sits in his own footnote rather than in anything added afterward.
What makes a sketch feel personal
The base effect is not the whole finding, and what has been added to it since 1949 matters more for understanding why a reading lands on one specific person. A review of the literature by D. H. Dickson and I. W. Kelly reports two moderators found repeatedly across the studies it surveys: a profile labelled as prepared for a particular person is accepted more readily than the identical profile labelled as a description of people in general, and favourable statements are accepted more readily than unfavourable ones 2.
Both conditions are satisfied, by construction, whenever the description in front of someone has been drawn from a birth date, time and place rather than issued as a stock character type. A natal chart reading meets the personalisation condition before a single word of interpretation is written, because it is addressed to one set of coordinates that belongs, near enough, to one person; whether it also meets the favourability condition depends on the practitioner producing it.
Gap in the evidence
Dickson and Kelly are candid about the state of the evidence behind these moderators: most of the component studies were run on university students, a limitation the review names itself and recommends addressing with research on the general public 2. Several individual findings, including whether the kind of test device a profile is said to come from changes how readily it is accepted, are reported as weak trends that do not reach significance in every replication. Whether the student-sample limitation has since been remedied by research on a broader population is not established in the sources consulted for this entry.
None of this is a finding about the person doing the rating. Forer’s own conclusions were addressed to clinicians rather than to patients: a sketch validated by the reaction of the person it describes has not thereby been shown to be an accurate instrument, because almost any sufficiently general and sufficiently generous sketch will produce that same reaction in almost anyone 1. What the effect explains is a fact about how self-description and generality interact in ordinary language, not a fact about which readers can be caught out and which cannot; the review’s own point is that the effect is close to universal 2.
That leaves room for what a competent practitioner is doing, which is not the recitation of a stock paragraph. On what has and has not been tested about chart interpretation itself, see Testing Astrology; on the newspaper format whose production method comes closest to Forer’s own identical sketch, see The Sun-Sign Column.
References
- Bertram R. Forer. The Fallacy of Personal Validation: A Classroom Demonstration of Gullibility. Journal of Abnormal and Social Psychology 44(1), 1949, 118-123. doi:10.1037/h0059240 recordpeer reviewed
The founding demonstration, read in full from a scan of the original journal pages rather than taken from the citation everyone repeats. Thirty-nine students in Forer's own introductory psychology class rated one identical thirteen-item sketch as a description of themselves. Two things the secondary literature gets wrong. The famous mean of "4.3" out of 5 conflates two different ratings: from Forer's own Table 1 the mean rating of the sketch works out at 4.26, and 4.31 is the mean rating of the diagnostic instrument, which is a different question. And the detail that the statements came largely from a newsstand astrology book is **Forer's own footnote**, not a later embellishment: it is on the page. That last point matters here, because it means the founding demonstration of the effect was built out of astrological material by the man who named the problem.
- D. H. Dickson, I. W. Kelly. The 'Barnum Effect' in Personality Assessment: A Review of the Literature. Psychological Reports 57(2), 1985, 367-382. doi:10.2466/pr0.1985.57.2.367 recordpeer reviewed
Read in full. The standard review of the effect and, more usefully, of its moderators: a profile labelled as prepared "for you" is accepted more readily than the same profile labelled generic, and favourable statements are accepted more readily than unfavourable ones. Those two findings explain more than the base effect does. Candid about its own limits, which is why it is cited rather than a livelier secondary account: most component studies used university students only, and several individual findings are weak or non-significant trends across replications. Do not present this literature as more settled than the review itself does.
Further reading
- C. R. Snyder, Randee J. Shenkel, Carol R. Lowery. Acceptance of Personality Interpretations: The 'Barnum Effect' and Beyond, 1977.