Machine Learning Predicts Echoic Verbal Behavior in Children with Autism Spectrum Disorder

Strategic Insights -

Researchers have made a significant breakthrough in predicting the likelihood of children with Autism Spectrum Disorder (ASD) developing echoic verbal behavior using machine learning strategies. By analyzing 143 case records of children assessed with the Verbal Behavior-Milestones Assessment and Placement Profile (VB-MAPP), a multilayer perceptron (MLP) model achieved 100% accuracy in predicting the presence or absence of an echoic repertoire. This study provides a general guide for using machine learning in conjunction with receiver operator characteristics analysis to predict outcomes in behavior analytic investigations.

Key Takeaways:

  • Researchers used a machine learning strategy to predict the likelihood of children with ASD developing echoic verbal behavior.
  • The study analyzed 143 case records of children assessed with the VB-MAPP and used a multilayer perceptron (MLP) model.
  • The MLP model achieved 100% accuracy in predicting the presence or absence of an echoic repertoire.
  • The study provides a general guide for using machine learning in conjunction with receiver operator characteristics analysis.
  • The research aims to provide more efficacious treatment for children with ASD by improving the prediction of outcomes in behavior analytic investigations.
  • The study highlights the potential of machine learning in predictions and implications for behavior analytic researchers.

Statistics:

  • 143 case records of children with ASD were analyzed in the study.
  • The MLP model achieved 100% accuracy in predicting the presence or absence of an echoic repertoire.
  • The study used a receiver operator characteristics analysis to evaluate the findings.

Sources:

  • osf.io: preprints/psyarxiv/zy7gm_v1/ - This preprint has not been peer-reviewed.