Words or Numbers? A Pilot Study of the Impact of Task Type on the Mathematics Performance of Autistic Children

Purpose: Discrepancies have previously been observed between the numerical mathematics problem solving of autistic children and their solving of worded mathematics problems, with the latter perceived as an area of challenge. While there are many reasons why this challenge may exist (e.g., reading comprehension and executive function differences), one, as yet untapped, explanation is the specific mathematical concepts assessed which may confound comparisons across assessments. Methods: In this study, 43 autistic children (32 male, 11 female) aged 6- to 12-years completed four assessments of mathematics problems: WIAT-III subscales of Numerical Operation and Mathematical Problem Solving, and two bespoke subscales (numerical and worded) matched on mathematical concepts and difficulty across numerical and worded problem sets. Results: We found that while performance in the worded problems subscale of the WIAT-III was significantly lower than the numerical problem subscale, there was no significant difference between the two task types in the bespoke measures. Conclusion: Our results suggest that differences in mathematical concepts assessed between worded and numerical problems may explain challenges experienced by autistic children in worded problems, rather than reading comprehension alone. As such, when drawing conclusions about the mathematics performance of autistic children, careful consideration of the mathematics concepts evaluated is needed. These findings also have implications for the development of formal school assessments and research on mathematics skills in autistic children.