Experiment II · protocol
The Forer Machine
Enter your birth details, read the description that follows, and rate how well it fits. Then find out what everyone else was reading.
This is a protocol, not a result
Collection has not opened, so there is no data on this page and no counter to watch. What is published here is the plan — filed in advance precisely so it cannot be adjusted once results start arriving. The charts below describe the null hypothesis, computed from the design itself.
The wager
People rate vague, universally applicable descriptions as strikingly accurate accounts of themselves — the Barnum effect, after Forer's 1949 classroom demonstration. Demonstrations of it are everywhere; controlled measurements are rarer, because a single generic text gives you a number with nothing to weigh it against.
So this register has two arms. Half of participants read a description built to fit nearly anyone: double-sided, mildly flattering, impossible to disprove. The other half read one of the same length and rhythm making concrete claims — about sleep, about animals, about numbers — that cannot be true of most people. If the first is rated substantially higher, the effect is real and measured rather than merely displayed.
Fixed in advance: Welch's two-sample t-test on mean rating, two-sided, α = .05, with Cohen's d and a Mann–Whitney U reported alongside because a 1–10 rating is ordinal. Reads at 1,000, 5,000 and 10,000 entries. Assignment is deterministic on your participant identifier, so reloading cannot move you between arms.
About the birth details
You will be asked for a birth date, time and place. That is the ritual, and it is theatre: the details stay in your browser and are never transmitted. No birth data reaches our server, our database, or the published dataset — there is no field in the request and no column in the table. Open your browser's network tab while you use this page and you will find nothing carrying it.
What is withheld until you have rated is only that the text is fixed and that a second version exists. Both are shown the moment you submit, on this same page.
Both descriptions, in full
You have a deep need for others to think well of you, though you are more critical of yourself than anyone watching would guess.
There is a difference between who you are in company and who you are alone, and you notice it more sharply than other people do.
Some of your hopes you have never said out loud, because saying them would make failing at them real.
You can be decisive when it genuinely matters, yet you turn small choices over far longer than they deserve.
You have learned to keep certain opinions to yourself in rooms where holding them would cost you something.
Praise lands oddly with you — welcome, and faintly suspect at the same time.
A stretch of your life that felt like drifting turned out to have been necessary, though you could not have known it while it lasted.
You are more comfortable being the person relied upon than the person doing the relying.
There is a version of your life you still think about occasionally, without regret exactly.
You dislike being managed, and you notice immediately when someone is attempting it gently.
You do your best work late at night, and you rarely accomplish anything of substance before the middle of the day.
You keep a written record of your days, and you reread earlier entries more often than most people would think to.
You are indifferent to animals, and other people's attachment to their pets strikes you as faintly puzzling.
Given a free evening, you would choose a crowded party over a long conversation with one close friend.
You think about the past very little; what is finished holds almost no interest for you.
You are physically restless, and sitting still through a long film is genuinely uncomfortable for you.
Numbers come to you more easily than words, and always have.
You have moved house often enough that no single place feels like the one you are from.
You decide quickly and almost never revisit a decision once it is made.
Silence in company makes you uneasy, and you are usually the one who fills it.
The filed protocol
The full preregistration is a document in the site's public repository. Its SHA-256 is printed below and computed from the file itself, so you can download it, hash it, and confirm that the plan described on this page is the plan that was filed. Every revision is a commit you can read.
SHA-256 2de1eed41f6ce65d3aa856b54c10bcc030ee5e3b412020a40980c9c65701e8d4
Known ways this could mislead
- The specific version might be rated low because it reads as oddly precise rather than because it fails to fit. Both texts are published in full so you can judge whether they are matched in length and register, which is the mitigation.
- A visitor to a site called Occult Research may guess a trick is coming and rate defensively. That biases toward finding no difference, which makes a positive result the more robust one.
- Any difference is a property of these two particular texts as much as of the categories they stand for. A replication with new texts is the correct way to test whether it generalises.
- A 1–10 rating is ordinal, not interval. That is why a rank-based test is reported beside the t-test rather than instead of it.