Criterion Validity, Scalability and Stability of Scoring on the Bayley-III in Children With Angelman Syndrome

Background: The Bayley Scales of Infant Development is used in many studies and clinical trials in children with developmental disabilities, including children with Angelman syndrome (AS). Method: We assessed 142 children with AS in an international multicentre study with the Bayley Scales of Infant Development III, of which 52 children were tested more than once. We assessed criterion validity using proportion analysis, scalability using Mokken analyses and stability of scoring by counting pass-to-fail and fail-to-pass items. Results: Results revealed good scalability in all scales but the expressive language scale, indicating that the items of these scales measured one underlying trait. In the expressive language scale, the AS-related speech difficulties invalidated scoring. Scoring within children across assessments was unstable for all scales except the gross motor scale, as more than half of the children made one or more errors in previously correct items. Loss or regression of skills does not fully explain this finding. Alternative explanations including motivation, concentration, on-task behaviour and anxiety should also be considered when scores decline. Conclusions: This study shows that caution should be taken when interpreting single and successive scores of children with AS on the Bayley-III and that other forms of assessment should complement assessment in children with AS.