How to talk to strangers when you have social anxiety

First, the boundary, because it matters more than anything else on this page. Social anxiety runs from "I dread this and do it anyway" to a diagnosable condition that shapes where somebody can work and live. Nothing here treats the second one. If avoidance is costing you work, friendships or study, that is worth taking to a doctor or a therapist, and it is a more effective route than any website.

What follows is for the milder end, and for the part of the harder end that is ordinary miscalibration sitting on top of something real.

The only finding here that involves doing it repeatedly

Most of this guide argues that a single wrong belief is the barrier. This is the one study that tested whether the belief can actually be moved.

Researchers ran a week-long scavenger hunt at two universities, one in the US and one in the UK. People in the treatment group were sent out each day to find, approach and talk to a stranger. The control group did nearly the same thing — find a stranger, approach them — but only observed them. Between them, the people who talked had over 1,300 conversations with strangers in a week.

Two things shifted, measured against the controls. They became less pessimistic about being rejected, and more confident in their own ability to hold a conversation. Both were still there when they were checked again a week after the study ended.

Why repetition and not advice

This fits the rest of the picture. If the barrier is a prediction that never gets tested, then reading that the prediction is wrong changes very little — you already believe things about yourself that arguments haven't shifted. What moves it is a run of actual outcomes, close enough together that they can't each be explained away as a fluke.

The researchers make roughly that point: occasional good conversations in ordinary life don't fix the pessimism, because a single conversation is easy to discount. Volume is what stops it being discountable.

What this means somewhere like here

Ours, not the research

The study was run on campuses, approaching people in person. Nobody has tested whether any of it transfers to typed conversation with strangers you cannot see, so the argument below is ours: an anonymous chat site looks to us like an unusually cheap place to run that experiment on yourself, for reasons that have nothing to do with courage.

Nothing is at stake socially. A conversation that goes badly is with somebody who does not know you and never will.
There is no approach. The part most people dread — the walking over, the timing, the visible interruption — does not exist in text.
You can leave. Which, on the numbers, matters more than it sounds: knowing the exit is cheap is what makes the entrance cheap.
Volume is available. The finding was about repetition, and repetition is the one thing that is easy here and hard everywhere else.

Being straight about the evidence

The study was run on university participants who signed up for a study, not on people selected for social anxiety, so strictly it says nothing about a clinical population. "At least a week" is exactly as long as the follow-up ran; nobody knows whether it holds for a year. And a real share of participants dropped out or failed an honesty check partway through — which is worth knowing both as a limit on the result and as a fact about the task itself. A week of approaching strangers is hard, and some people who signed up for it stopped.

So: promising, one study, short follow-up, wrong population to draw conclusions about the worst cases from. That is a reasonable basis for trying something cheap and a poor basis for anything else.

Cheap is the operative word, and it is the only claim being made for a site like this one: a conversation here costs an access code and can be left the moment you want to leave it. That is a fair thing to try a few times and find out. It is not a course of treatment, and if the avoidance is heavy enough to be shaping where you work or whether you go out, that still belongs with a doctor — trying the cheap thing first is not a reason to put the other off.

Where this comes from

A week of repeatedly talking to strangers made people less pessimistic about being rejected and more confident in their ability to hold a conversation, and the shift was still there a week later.

What the study actually did

A week-long intervention run at two universities, one in the US (135 people) and one in the UK (151). A phone app set daily scavenger-hunt missions. The treatment group had to find, approach and talk to a stranger; the control group found and approached a stranger but only observed them. The treatment group had more than 1,300 conversations between them over the week. Both fear of rejection and belief in one's own conversational ability had moved relative to controls, and were re-measured a week after the study finished.

What it doesn't show. Anything about clinical social anxiety. Participants were university study volunteers, not people selected or diagnosed for an anxiety disorder, so this is not evidence about treatment. "Lasting" means as long as the follow-up ran — one week. Between 12% and 17% of participants dropped out before four days and a further 11% to 18% failed an honesty check, which limits the result and also says something about how hard the task was. One study, not yet independently replicated.

Sandstrom, G. M., Boothby, E. J., & Cooney, G. (2022). Talking to strangers: A week-long intervention reduces psychological barriers to social connection. Journal of Experimental Social Psychology, 102, 104356 · 10.1016/j.jesp.2022.104356