Published on 12, July, 2020
I've just done one and my result was 39. Apparently the border is 31. How accurate are these tests?
Generally speaking the AQ50 (assuming that's what you did?) has a sensitivity in the range of about 85% or a bit better, in other words if 100 people with autism take it, 85 of them will score above the threshold & be "identified" by the test. Unfortunately, the AQ50 doesn't have high enough "specificity" to be taken as conclusive - in other words, if 100 people *without* autism take it, a significant number of them will also score above the threshold (see here for e.g. www.ncbi.nlm.nih.gov/.../). This is why diagnosis requires a whole host of methods deployed by specifically-trained professionals.
If you take some other *different* tests as well, you can form some better conclusions maybe - e.g. the Aspie Test (rdos.net), EQ50, RAADS-R......