# Item Analysis of an Early Social Responsiveness Scale for Assessing Autism Risk

## TL;DR

This study analyzes items from a scale used to detect autism risk in young children to identify which items best distinguish autism risk from non-risk.

## Contribution

The study provides the first item analysis of the Early Social Responsiveness Scale to optimize its use in identifying autism risk.

## Key findings

- Ease of social engagement items were most effective at differentiating ASD risk.
- Hat and tickle activities were least effective in indicating ASD risk.
- The analysis could help improve the clinical utility of the ESR assessment.

## Abstract

Early diagnosis of autism spectrum disorder (ASD) is vital for effective intervention and improves social and behavioral development. The previous literature has shown that the Early Social Responsiveness (ESR) assessment is effective at detecting ASD risk in individuals as early as 13 months of age (“parent study”). However, an item analysis that examines individual item scores has not been conducted to further elucidate the strength of this assessment. In this study, we analyzed an existing dataset (collected in the parent study) containing individual item responses from the ESR assessment of 120 children (n = 61 males and n = 59 females; age range = 15–24 months). Through item analysis, we determined which ESR items or item sets are best at differentiating ASD risk from non-ASD risk. Ease of social engagement (i.e., questions assessing the administrator’s perceived level of effort in engaging the child) was the most effective risk indicator, with the hat and tickle activities being least effective at indicating ASD risk. These results could contribute to optimizing the scale and facilitating its clinical adoption.