Sensitivity and Expected Change of Commonly Used Social Communication Measures in Longitudinal Research of Young Autistic Children.
Standard social-communication tests grow slowly in young autistic kids, so plan for small wins and choose sensitive measures.
01Research in Context
What this study did
Sterrett et al. (2025) pooled every long-term study that tracked social-communication scores in autistic children under six.
They asked one question: how much do the usual tests actually move as kids grow?
What they found
Most scales barely budge. Small to medium gains are the best you can expect.
In plain words, a child who starts at 30 months with few words will still score low at 48 months, even if he now asks for cookies.
How this fits with other research
Cunningham (2012) warned us first: no single tool captures early social change. Kyle’s numbers now prove that warning was spot-on.
Barnard-Brak et al. (2016) and Jones et al. (2007) show the SCQ can swing from great to poor accuracy depending on age. Kyle folds those same SCQ papers into the meta and still finds tiny change over time—so the test’s shaky sensitivity is part of the problem, not just who you screen.
Gillespie-Lynch et al. (2019) meta-analyzed links between social skills and skills like theory of mind. They also found only small links. Kyle’s result lines up: if the building blocks barely shift, the overall social score won’t either.
Why it matters
Stop waiting for big jumps on your favorite checklist. Pick tools that show even small steps—like teacher-friendly play scales from Shire et al. (2018)—and celebrate micro-gains instead of moving the goal post.
Get CEUs on This Topic — Free
The ABA Clubhouse has 60+ on-demand CEUs including ethics, supervision, and clinical topics like this one. Plus a new live CEU every Wednesday.
Graph the last three months of each child’s social scores; if you see a flat line, add a play-based probe that records new joint-attention bids even if the formal scale stays put.
02At a glance
03Original abstract
Social communication measures used to track change in autistic children have not undergone rigorous psychometric evaluation. There is little data on their expected change or sensitivity to change. Meta-analytic techniques were used to examine sensitivity to change and expected change over time and whether these are influenced by factors like children's age and the presence of intervention. Pooled effect sizes were generated within measures, rather than within broader constructs. Change over time was small to medium, although there was variability. Most outcomes were not sensitive to change over time. Change in some measures was influenced by child characteristics and methodological characteristics of included studies such as study quality and the method of scoring measures (e.g., using age-equivalents versus standard scores). Tests measuring similar constructs can vary in their expected change, and so care is needed when selecting them.
Journal of autism and developmental disorders, 2025 · doi:10.1097/01.chi.0000196423.80717.32