Skip to content
How it worksQuestionsKnowledgeContact
Log inde
Knowledge

Research

The 36 questions — what Aron’s experiment proves and what it does not

Aron’s famous experiment in detail: setup, result, limits — and how to use the principle as a couple without just working through the list.

8 min read

AI-generated

Few psychology experiments have been retold so often and cited correctly so rarely. The "36 questions that lead to love" come from a 1997 study by Arthur Aron and colleagues with the dry title "The Experimental Generation of Interpersonal Closeness". It became famous eighteen years later through a newspaper article — and in the process almost everything that makes it interesting got lost.

The actual setup

Students who did not know each other were paired up and given 45 minutes. One group worked through three sets of twelve questions that got more personal from set to set. The control group got tasks about as personal as a conversation in a queue. Afterwards, felt closeness was measured.

The question group reported markedly more closeness. In one variant of the experiment a pair did later marry — the anecdote that gave the whole thing its name. The study itself, however, never examined romantic love. It examined how quickly closeness can be produced between any two people.

What the study does not show

  • It does not show that 36 questions create love. It shows they create closeness — two different things.
  • It says nothing about durability. Closeness was measured immediately afterwards, not a year later.
  • It says nothing about existing couples. Participants were strangers; for long-term relationships the finding applies only by analogy.
  • The sample was small and student-based. Like most work from that era, the study is a pointer rather than a proof.

None of which makes it worthless — quite the opposite. It only shifts what you should take away: not the list, but the principle.

Applying the principle to an existing couple

For strangers the escalation works in 45 minutes because it starts from zero. Couples together for years have a different problem: they believe they already know the answers. Research on the "illusion of insight" shows long-term couples often predict their partner’s preferences worse than strangers do — they stop asking because they feel certain.

So the application changes. Instead of 45 minutes in one sitting: a long arc in small steps. Instead of a fixed list: questions that fit the phase you are actually in. And instead of "What do I know about you?", the question becomes "How much of what I know about you still holds today?"

Three questions from the set that rebuilds the same trajectory for couples — light, personal, close:

  • What did you last read, hear or watch that stayed with you?
  • What did you dream about when you were seventeen?
  • What do you think I most often misread about you?

Four minutes of eye contact

The original includes an often-forgotten ending: four minutes of silent eye contact. Popular retellings tend to dismiss it as kitsch; in the experiment it was part of the procedure. Try it and you quickly see why: four minutes is uncomfortably long, and that is exactly the point. The mechanism is the same as with the questions — showing yourself and staying there.

Sources

  1. Aron, A., Melinat, E., Aron, E. N., Vallone, R. D. & Bator, R. J. (1997). The Experimental Generation of Interpersonal Closeness. Personality and Social Psychology Bulletin, 23(4), 363–377.
  2. Sprecher, S., Treger, S., Wondra, J. D., Hilaire, N. & Wallpe, K. (2013). Taking Turns: Reciprocal Self-Disclosure Promotes Liking. Journal of Experimental Social Psychology, 49(5), 860–866.
  3. Swann, W. B. & Gill, M. J. (1997). Confidence and Accuracy in Person Perception. Journal of Personality and Social Psychology, 73(4), 747–757.

Common questions

Can a couple simply go through the 36 questions?

You can, but it is rarely the best route. The list was written for strangers; some questions are trivial for couples, others skip intermediate steps. A set that uses the same escalation but assumes an existing relationship works better.

How long does the effect last?

That is an open question. Closeness was measured immediately afterwards. Anything that lasts weeks comes from repetition, not from the single session.

Knowledge