Can structured self-disclosure create closeness?

Structured self-disclosure can create a stronger immediate feeling of closeness between strangers. That is the result behind the famous 36-question procedure.

It is also narrower than the internet version. The original researchers built a laboratory procedure for producing temporary felt closeness. They did not test a dating formula, promise that two people would fall in love or show that one conversation creates loyalty, dependence, commitment or a durable relationship.

The original comparison was disclosure versus small talk

In the 1997 experiment, stranger pairs worked through three 15-minute sets of tasks. The prompts became progressively more personal and required reciprocal answers or other relationship-building actions. A comparison group spent the same time on structured small-talk prompts with minimal personal disclosure. Pairs were not required to complete every slip. The experiment tested the timed task bundle, not all 36 prompts as individually necessary ingredients.

The study included 50 pairs of psychology students: 33 mixed-sex pairs and 17 pairs of women. After 45 minutes, the closeness-condition mean was 4.06 on the authors’ composite measure, compared with 3.25 for small talk. The reported standardized difference was d = 0.88.

That is strong evidence for a difference inside this experiment. It is not an estimate of what will happen to every pair outside the classroom. Participants knew they were in a study, the sample was narrow and the outcome was reported immediately after the interaction.

The task bundle mattered more than simply spending 45 minutes together

The result was not produced by putting two strangers into any structured conversation. The control pairs also talked for 45 minutes. The distinguishing condition combined escalating, reciprocal and personal disclosure with intimacy-related tasks.

Across the authors’ next two studies, three other ingredients did not show a significant overall closeness effect: matching pairs so they did not disagree on important attitudes, telling them to expect mutual liking, and explicitly making closeness the goal.

Those nulls do not prove the ingredients never matter. They show that the evidence for the procedure rests most clearly on the nature of the exchange, not compatibility screening, positive expectation or being told to connect.

The experiment created a feeling, not a complete relationship

The authors were explicit about the boundary. Their goal was temporary felt closeness—an immediate sense of connection—not an ongoing relationship.

Across the closeness conditions in all three studies, the average score on one measure was 3.82. In a separate student sample, the mean rating for a person’s closest relationship was 4.65. The procedure moved strangers meaningfully, but it did not make a new interaction equivalent to an established bond.

A seven-week follow-up in Study 3 found that some pairs talked, sat together or did something together again. There was no control group for that follow-up, so the researchers could not say how much of that contact the procedure caused.

Newer studies support disclosure while narrowing the claim

The broader evidence is compatible with the original result, but it does not turn one script into a universal formula.

A 2021 randomized study assigned 70 new-acquaintance dyads to high- or low-disclosure questions. One dyad was excluded a priori and 63 had usable attunement scores. High-disclosure pairs revealed more personal information. Within that reduced analytic sample, greater observed disclosure was associated with more similar cortisol changes within the pair. That physiological attunement is not the same as a durable relationship or a direct closeness effect.

Another 2021 study randomized 105 dating or married couples to higher- or lower-closeness discussions. The higher-closeness group reported only marginally more disclosure and marginally more closeness. It showed larger cortisol and testosterone increases, but self-reported closeness did not explain those hormonal changes.

A 2026 speaker-listener study compared emotional self-disclosure with non-disclosure while measuring both partners’ brain activity. It analyzed 40 same-sex dyads of existing student friends after excluding two of 42 recruited dyads. The disclosure group reported greater closeness. Interpersonal neural synchrony was associated with closeness and entered a modeled mediation pathway; empathy and gender were also associated with parts of the pattern. This adds a newer neural correlate, not evidence that a specific list of questions works for every person or relationship.

Depth alone is not the mechanism

The original procedure was reciprocal: both people answered, the turns alternated and intensity rose gradually. It did not ask one person to expose themselves while the other extracted information.

The 2026 study likewise points toward a speaker-listener process in which affective and cognitive empathy were associated with closeness. It measured the listener’s brain activity, not responsive listener behavior, and therefore does not show that any observable response caused the closeness difference.

This is why ask deeper questions is incomplete advice. A revealing prompt can increase vulnerability without producing safety, care or mutuality. The studies do not authorize pressure, surprise disclosure or treating another person’s boundaries as resistance to be overcome.

What the 36-question procedure can and cannot do

The procedure is best understood as a structured opportunity for mutual attention and graduated disclosure. It can change an interaction. It cannot decide what the relationship should become.

Original sources

• Aron et al., 1997 original procedure
• Thorson et al., 2021 new-acquaintance study
• Chin et al., 2021 couples study
• Yu et al., 2026 speaker-listener study

Independent editorial summary. The authors are not affiliated with LifeScore.