The six ratings
- Many good trials agree, including trials against a fair comparison such as a placebo or another treatment.
- Several trials show it helps, but fewer of them, or mostly against waiting lists.
- A few small trials show it helps. Bigger or fairer trials are still needed.
- Promising early results, very small trials, or trials only in a lab.
- No trial has tested it for social anxiety. People may still report that it helps them.
- Trials tested it for social anxiety and it did not help more than the comparison.
Extra labels
- Results disagree
- Good studies point in different directions.
- Best inside therapy
- The evidence is for using it as part of therapy, not on its own.
How a rating is decided
- We look for trials in people with social anxiety. Results for other anxiety problems count only as a hint.
- We prefer studies that combine many trials. One small trial can be lucky. Many trials that agree are harder to argue with.
- We ask what it was compared with. Beating a waiting list is easier than beating a placebo or another treatment.
- We check who ran the trial. When the company that sells a product ran the trial, the page says so.
- We say what people report too. What people say on forums is not evidence that something works. It does tell you what to expect, and pages show it next to the trials.
- Every page is checked again at least every 12 months. The date at the top says when it was last checked against its sources.
Why fair comparisons matter
Many treatments look good in small trials against a waiting list. People on a waiting list know they are not getting help yet, and they often feel no better. So almost any active treatment beats them.
Results tend to shrink as trials get bigger and fairer. This happened with virtual reality exposure, apps and attention-training programs. That is why a rating looks at the fairest comparison available, not the biggest number.