TY - JOUR
T1 - Gender bias in autism screening
T2 - Measurement invariance of different model frameworks of the Autism Spectrum Quotient
AU - Belcher, Hannah L.
AU - Uglik-Marucha, Nora
AU - Vitoratou, Silia
AU - Ford, Ruth M.
AU - Morein-Zamir, Sharon
N1 - Funding Information:
N.U.-M. and S.V. received funding from the National Institute for Health and Care Research (NIHR) Maudsley Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London. N.U.-M. was funded by NIHR (Pre-doctoral Fellowship (NIHR301151); Doctoral Fellowship (NIHR302618)). The views expressed are those of the author(s) and not necessarily those of the National Health Service, NIHR or Department of Health and Social Care.
Publisher Copyright:
Copyright © The Author(s), 2023. Published by Cambridge University Press on behalf of the Royal College of Psychiatrists.
PY - 2023/10/2
Y1 - 2023/10/2
N2 - Background The Autism Spectrum Quotient is a popular autism screening tool recommended for identifying potential cases of autism. However, many women with autism demonstrate a different presentation of traits to those currently captured by screening measures and assessment methods, such as the Autism Spectrum Quotient. Aims Different models of the Autism Spectrum Quotient have been proposed in the literature, utilising different items from the original 50-item scale. Within good-fitting models, the current study aimed to explore whether these items assess autistic traits similarly across men and women. Method Seventeen Autism Spectrum Quotient models were identified from the literature. Using the responses of a large sample of adults from the UK general population (5246 women, 1830 men), confirmatory factor analysis was used to evaluate the fit of each model. Measurement invariance with respect to gender, adjusting for age, was explored in the 11 model frameworks that were found to have satisfactory fit to our data. Results It emerged that only two items were gender invariant (non-biased), whereas for the remaining items, the probability of endorsement was influenced by gender. In particular, women had a higher probability of endorsing items relating to social skills and communication. Conclusions If the items of the Autism Spectrum Quotient indeed reflect autism-related traits, those items should be rephrased to ensure they do not present a gender-related bias. This is vital for ensuring more timely diagnoses and support for all people with autism.
AB - Background The Autism Spectrum Quotient is a popular autism screening tool recommended for identifying potential cases of autism. However, many women with autism demonstrate a different presentation of traits to those currently captured by screening measures and assessment methods, such as the Autism Spectrum Quotient. Aims Different models of the Autism Spectrum Quotient have been proposed in the literature, utilising different items from the original 50-item scale. Within good-fitting models, the current study aimed to explore whether these items assess autistic traits similarly across men and women. Method Seventeen Autism Spectrum Quotient models were identified from the literature. Using the responses of a large sample of adults from the UK general population (5246 women, 1830 men), confirmatory factor analysis was used to evaluate the fit of each model. Measurement invariance with respect to gender, adjusting for age, was explored in the 11 model frameworks that were found to have satisfactory fit to our data. Results It emerged that only two items were gender invariant (non-biased), whereas for the remaining items, the probability of endorsement was influenced by gender. In particular, women had a higher probability of endorsing items relating to social skills and communication. Conclusions If the items of the Autism Spectrum Quotient indeed reflect autism-related traits, those items should be rephrased to ensure they do not present a gender-related bias. This is vital for ensuring more timely diagnoses and support for all people with autism.
KW - Autistic spectrum disorders
KW - community mental health teams
KW - neurodevelopmental disorders
KW - psychological testing
KW - statistical methodology
UR - http://www.scopus.com/inward/record.url?scp=85174178360&partnerID=8YFLogxK
U2 - 10.1192/bjo.2023.562
DO - 10.1192/bjo.2023.562
M3 - Article
AN - SCOPUS:85174178360
SN - 2056-4724
VL - 9
JO - BJPsych Open
JF - BJPsych Open
IS - 5
M1 - e173
ER -