Early detection of Autism Spectrum Disorder (ASD) can improve outcomes for children, yet the effectiveness and validity of universal screening methods has been questioned. Now, researchers have created a new framework to generate a valid early ASD screening method using a novel approach based on “face and content validity”. Content validity measures the actual extent to which a test measures all aspects of what it claims to measure (in this case signs of early autism), whereas face validity assesses the superficial level by which a test appears to measure the underlying item. A network of European experts first identified what they considered to be the main factors involved in early ASD. 12 psychological constructs (most notably “social interaction”, “interest in others” and “joint attention”) and eight corresponding test items specific for early ASD (ages 14-36 months) were selected with the highest level of consensus in the group. These parameters were then combined to derive a new theoretical model, based on the DSM-5 diagnostic criteria for ASD, which achieved good face and content validity to diagnose early ASD. The researchers hope that their model will serve as a theoretical framework to develop effective and valid screening tests for early ASD that practitioners feel confident to use.
Magán-Maganto, M. et al. (2017), Measurement Issues: Building a theoretical framework for autism spectrum disorders screening instruments in Europe. Child Adolesc Ment Health, doi:10.1111/camh.12256
Glossary
Content validity: the extent by which a test measures all aspects of what it claims to measure
Constructs: a tool to help understand human behaviours that are not directly observable, such as cognitions, emotions or attitudes
DSM-5: fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (2013), detailing the criteria to evaluate individuals for autism spectrum disorder and other social communication disorders
Face validity: the superficial extent by which a test appears to be able to measure what it claims it will measure