Do the 36 Questions Work? What Aron's Study Found

The 36 questions have had a long second life as a viral dating experiment. The story usually goes: two strangers ask each other 36 questions, stare into each other's eyes, and fall in love. The real study is more modest and more useful. It's one of the clearest experimental demonstrations of how closeness forms, and it says a lot about what makes a first conversation go somewhere.

Where the 36 Questions Come From

The questions are the appendix to "The Experimental Generation of Interpersonal Closeness: A Procedure and Some Preliminary Findings," by Arthur Aron, Edward Melinat, Elaine N. Aron, Robert Darrin Vallone and Renee J. Bator, published in Personality and Social Psychology Bulletin in April 1997.

The goal was methodological. Relationship researchers could compare people who were already close, but they couldn't create closeness in a lab and test what caused it. Aron's team wanted a procedure that would, in their words, develop "a temporary feeling of closeness" between strangers under controlled conditions. Romance wasn't the point.

The procedure was run with undergraduate psychology students during a class session. Students were paired with someone they didn't know, either a man and woman or two women. Each pair got three envelopes of slips. Every 15 minutes, the experimenter told them to move to the next set. The questions got more personal as they went.

What Are the 36 Questions?

The three sets escalate deliberately.

Set I starts light: "Given the choice of anyone in the world, whom would you want as a dinner guest?" and "What would constitute a 'perfect' day for you?" It ends with tasks like "Take 4 minutes and tell your partner your life story in as much detail as possible."

Set II moves to values and memories: "What is your most treasured memory?", "What is your most terrible memory?" and "How do you feel about your relationship with your mother?"

Set III asks for direct vulnerability with the other person: "Tell your partner what you like about them; be very honest this time," "When did you last cry in front of another person?" and, last, "Share a personal problem and ask your partner's advice on how he or she might handle it."

Two details get lost in most retellings. First, both people answer every question, alternating who goes first. The exchange is the design. Second, pairs weren't required to finish all 36. The instructions told them to take their time with each one. There's also no eye-gazing exercise in the published procedure; that was added later in popular versions.

What the Study Actually Found

The paper reports three studies.

Study 1 tested the questions against small talk. A comparison group got slips with low-disclosure prompts like "Describe the last time you went to the zoo." On a combined closeness measure, the question pairs averaged 4.06 and the small-talk pairs 3.25. The authors calculated that difference at about 0.88 standard deviations, which they note is large by the standards of psychology research. The content of the conversation, not just time spent together, drove the effect.

Study 2 tested matching and expectations. Some pairs were matched so they didn't disagree on attitudes that were important to them, and some were told to expect their partner would like them. Neither made a significant difference to closeness.

Study 3 tested whether saying "get close" mattered. Some pairs were told closeness was the goal; others weren't. Again, no significant overall difference, which suggests the effect isn't just people doing what they think the experimenter wants.

How Close Did People Get?

The authors compared participants' scores to a separate sample of students at the same university who rated their "closest, deepest, most involved, and most intimate relationship." After about 45 minutes with a stranger, participants' average closeness rating was around the 30th percentile of those ratings. Put another way, the new relationship was rated as closer than the closest relationship in the lives of about 30% of similar students.

That's striking, and the authors flag the caveat themselves: people may use the scale differently when rating someone they just met in an experiment.

In Study 3, the researchers also ran a follow-up questionnaire about seven weeks later. Of the 58 pairs represented:

• 57% had had at least one more conversation
• 35% had done something together
• 37% had sat together in class

There was no control group for that follow-up, so the authors note it's possible that much contact would have happened anyway.

Did Anyone Fall in Love?

The famous detail is that a couple from the study got married. That's real but anecdotal. The paper mentions, in passing, that an earlier version of the procedure (run for about an hour and a half with mixed-sex stranger pairs) produced reports of lasting impact, "including one pair who married!" It was never a measured outcome.

The authors were direct about the limits. Asking "are we producing real closeness?", they answered "Yes and no." The feeling of closeness seemed similar in important ways to closeness that develops naturally, but they wrote that it seems unlikely the procedure produces "loyalty, dependence, commitment, or other relationship aspects that might take longer to develop."

Two Findings Worth Knowing About

Because the researchers could control who got paired with whom, they also ran some exploratory analyses.

Attachment style mattered. Pairs involving avoidant partners, particularly the type the authors call avoidant/dismissive, reported less closeness. That fits a large body of correlational research on attachment styles in relationships, and the authors argue their experimental setup supports a causal direction from attachment style to relationship quality. They describe the findings as tentative.

Introverts weren't worse at getting close. Introverts reported somewhat less closeness overall, but when closeness was made an explicit task, introverts became as close as extraverts. The authors suggest introverts may be uncomfortable with unstructured stranger interaction rather than unskilled at intimacy. If that sounds familiar, our guide to dating in NYC as an introvert builds on the same idea.

What This Means for a First Date

You don't need to bring 36 slips of paper to dinner. The study's mechanism transfers without the script.

Escalate gradually. The questions start with dinner guests and perfect days and only get to family and regrets after 30 minutes. Opening with your most personal story feels like oversharing; reaching it naturally feels like connection.

Make it mutual. Disclosure works when it's reciprocal. If one person answers everything and the other interviews, it doesn't produce the same effect. This is a recurring theme in the question that predicts a second date.

Content beats chemistry checklists. The small-talk group spent the same 45 minutes together and felt less close. Being in a room isn't enough; what you talk about matters. That fits what speed-dating research suggests about how poorly people's on-paper preferences predict who they click with in person.

Short, structured formats can work. A big takeaway is how much happened in under an hour. Structure helped strangers get past the opening awkwardness quickly. It's part of the case for shorter first dates: a short conversation with real content tells you more than a long one spent on logistics.

The Bottom Line

The 36 questions don't make people fall in love, and the researchers never said they did. What the study showed is narrower and more practical: gradual, mutual self-disclosure creates a real feeling of closeness between strangers, quickly, and more reliably than being matched on attitudes or told you'll like each other. Whether that becomes something more depends on everything the study didn't measure.

Sources:

• The Experimental Generation of Interpersonal Closeness: A Procedure and Some Preliminary Findings — Arthur Aron, Edward Melinat, Elaine N. Aron, Robert Darrin Vallone and Renee J. Bator, Personality and Social Psychology Bulletin 23(4), 363–377 (1997)