What Aron's 36 Questions Study Actually Showed
In 1997, Arthur Aron, Edward Melinat, Elaine Aron, Robert Vallone and Renee Bator published a procedure in Personality and Social Psychology Bulletin that has been quoted, misquoted and turned into a listicle ever since. Two strangers sat down together and worked through three sets of questions that grew steadily more personal, taking turns to answer every one. The session ran about 45 minutes. Afterwards the pairs reported feeling markedly closer to each other than strangers who had spent the same 45 minutes on small talk.
It is worth being precise about what that study did and did not show, because the popular version has drifted a long way from the paper. It showed that escalating, mutual self-disclosure — both people answering, each answer a little more personal than the last — generates closeness between strangers, and that it can happen surprisingly fast. It did not show that anyone falls in love in a handful of minutes, and it was never a matching algorithm: nobody was paired by their answers, and the questions were not a compatibility test. Mutual is the load-bearing word in the whole finding.
That is the real reason questions are interesting for dating, and it is a narrower reason than the internet usually gives. What the research supports is that asking and answering is how two strangers begin to like each other. Whether the person you come to like turns out to be a good partner is a separate question that no set of prompts can answer. What Qulo takes from this is the sequence, not a prediction: the exchange happens first, and the face comes after.
Reciprocity Is the Part That Does the Work
Sprecher, Treger, Wondra, Hilaire and Wallpe tested exactly this in the Journal of Experimental Social Psychology in 2013. Pairs of strangers either took turns disclosing — both asking, both answering — or split the roles, one person talking while the other listened, and then swapped. Taking turns produced more liking and more closeness. The same amount of information passed between the same two people; only the shape of the exchange changed.
Huang, Yeomans, Brooks, Minson and Gino found the dating-relevant version of that in the Journal of Personality and Social Psychology in September 2017. Across 110 speed daters and 1,961 second-date decisions, the people who asked more questions — follow-up questions above all, the kind that show you actually heard the previous answer — were more likely to be asked back. Not the most attractive, not the wittiest. The ones who asked.
A photo and two lines of bio cannot do any of that. They travel in one direction, they ask nothing of you, and there is nothing there to respond to. A question is the smallest object that makes both people put something down.
Attachment Theory: A Framework, Not a Verdict
John Bowlby's attachment theory is the standard vocabulary for how people behave in close relationships — secure, anxious, avoidant — and it is a real framework, worth naming as one. It is also worth saying what it is not. Bowlby was writing about children and caregivers decades before anyone swiped on anything, and nobody has shown that a dating app moves an adult's attachment style in either direction. Anyone telling you an app can fix or ruin how you attach is selling something.
What the design of an app can change is narrower and more mundane: what a no means. Where matching runs on photos, not matching reads as a verdict on your face, because your face was the whole of what was on offer. Where matching runs through someone's questions, a miss reads as a miss — you did not get the answers right. That is a difference in framing, not a therapeutic claim, and framing is the honest size of it.
Cognitive Compatibility and Sternberg's Triangle
Cognitive compatibility is a plain idea in an academic coat: two people whose thinking runs along similar lines — how they approach a problem, what they take for granted, what they find funny. Questions surface it early, because the way somebody frames a question tells you as much as the answer does. What nobody has shown, and what this article will not claim, is that getting a stranger's questions right predicts a relationship. It predicts that you paid attention.
Robert Sternberg's triangular theory of love is useful here as a description rather than as evidence. It splits lasting love into three components: intimacy, passion and commitment. A photograph speaks to one corner of that triangle at most, and it speaks to it before you know anything else. The other two corners get built in conversation, if they get built at all. That is a way of thinking about what a photo-first first impression leaves out — not a measurement of anything, and no number belongs to it.
Why a Photo Tells You Less Than It Seems To
None of this makes attraction unimportant. It makes a photo a poor container for it. Four familiar problems, none of which need a statistic to be obvious:
- The halo effect: we quietly assume that someone we find attractive is also kinder, funnier and more competent. It is one of the oldest documented biases in social psychology, and it is wrong roughly as often as it is right.
- A photo is chosen: it is the best of a few hundred, taken in good light, at the angle that works, on a day that went well. An honest picture of a face and an unreliable one about everything else.
- Looks stop being new information: the first impression is most of what a face has to tell you, and after that it repeats. What someone is like to talk to does not run out that way.
- A photo cannot answer: nothing you learn from it arrived in response to anything you offered — which is precisely the reciprocity the studies above say matters.
That is a limit on what photo-first matching can tell you. It is not evidence that questions predict who you should be with. Only that a photo was never going to, and that an exchange leaves you with something to go on that a face cannot give you.
What Qulo Actually Does With This
Qulo does not claim to run scientifically validated matching, because no such thing exists in dating. What it does is put the exchange first, mechanically, so that it cannot be skipped:
- You write 2 to 4 questions about yourself — multiple choice, four options each — and mark the option that is true of you. A paid plan raises the ceiling to 10.
- Someone matches by getting every one right. A single wrong answer means no match, so nobody arrives by tapping through.
- The questions come before the face, which reverses the usual order of a first impression.
- Solving takes attention. Effort is a weak guarantee of anything, but it is real evidence that the other person read what you wrote.
- The optional powers — a hint, halving the options, skipping a question — cost in-app diamonds and are never required to match. Every question is solvable without spending anything.
How to Write Questions Worth Answering
Because your answer has to be findable, writing good questions is a specific skill. The work above suggests what to aim for: personal enough to be worth disclosing, and answerable by someone paying attention rather than guessing.
- Be specific about yourself: "the city I would move to tomorrow" beats "my favourite city". One is a fact about you; the other is small talk.
- Make it findable, not psychic: the right answer should be reachable from your profile, your photos and a little thought. A question only your sister could answer is not a filter, it is a wall.
- Avoid decorative options: four plausible options with one true one is a question. One true option and three jokes is a formality.
- Ask about choices, not trivia: what you would do, what you would give up, where you would go. Choices show a person; facts about them mostly do not.
- Vary the depth: open with something light and put the one that actually matters to you third or fourth. That escalation is the part the 1997 study was really about.
The shortest route to knowing someone is a question you both have to answer. That is not a discovery — it is just what a conversation is.
The Honest Version
Question-based matching is not a proven route to a relationship, and this article is not going to pretend otherwise. What the research supports is smaller and still worth building on: asking questions, and answering them in turn, is how strangers begin to like each other — Aron and colleagues in 1997, Sprecher and colleagues in 2013, Huang and colleagues in 2017 all point the same way. A photo is not that. Qulo's only real claim is a claim about design: it makes the exchange the entry point instead of an afterthought, and leaves it to you to decide what the answers mean.