“These questions are geared towards male autistics, not female… I found they never quite reflected my experience”: Assessing the Efficacy of Autism Questionnaires in Identifying Adult Autistic Women

Traditional measures of autistic traits such as the Autism Quotient (AQ) are argued to be male biased and may lead to missed or mis-diagnosed autism among females. The current study examined the efficacy of three questionnaires to identify adult autistic women: Autism Quotient (AQ50; male bias); Girls Questionnaire-Autism Spectrum Condition (GQ-ASC; modified for adult women); Camouflaging Autistic Traits Questionnaire (CAT-Q; gender neutral). A sample of 343 autistic and 345 non-autistic women completed each questionnaire and the proportion of those scoring above or below each suggested cut-off were used to analyse specificity and sensitivity. The AQ50 was less sensitive in correctly identifying autistic women than both the CAT-Q and GQ-ASC. However, the AQ50 showed high specificity in correctly identifying most non-autistic women. The CAT-Q and GQ-ASC were equally highly effective in identifying autistic women but over half of non-autistic women scored above the cut-off for the GQ-ASC making it the least specific. Additional evaluative ratings and comments showed that autistic women found the AQ50 to be stereotyped and the least clear to interpret. The GQ-ASC was overall preferred despite some questions not feeling age appropriate. This study highlights the need for assessment questions that reflect the broader diversity of autistic traits.