A Concurrent Validity Study of the Mullen Scales of Early Learning (MSEL) and the MacArthur-Bates Communicative Developmental Inventory (CDI) in Infants with an Elevated Likelihood or Diagnosis of Autism.
Mullen and CDI tap different language skills in infants—don’t swap their scores when tracking early vocabulary in autism-risk cases.
01Research in Context
What this study did
The team gave two quick language checks to 14-month-old babies. Some babies had an older sibling with autism, so they were at higher risk. Others had no family history.
Each baby sat for the Mullen Scales, a short play test run by a trained adult. Parents also filled out the MacArthur-Bates CDI, a checklist of words and gestures their child uses at home.
What they found
The two tools only matched moderately. Babies later diagnosed with autism scored lower on both, but the gap size changed depending on which test you read.
In other words, a low Mullen score did not always line up with a low CDI score for the same child.
How this fits with other research
La Valle et al. (2025) saw the opposite pattern in toddlers with Down syndrome. In their sample, Mullen and parent-report Vineland-3 age scores lined up closely. The difference is likely age and diagnosis: toddlers with Down syndrome may show more stable skills than 14-month-olds with early autism signs.
Morrison et al. (2017) also found parent report and direct testing mostly agreed in 2-year-olds across autism, delays, and typical groups. Their null result seems to clash with the weak link seen here, but they used a generic parent form, not the CDI, and tested older children. Small changes in tool and age can shift agreement.
Sterrett et al. (2025) warn that most social-communication measures barely pick up growth over time in young autistic children. Their meta-analysis includes CDI data and supports the current finding: single low scores should be read cautiously and never alone.
Why it matters
When you assess an infant at risk for autism, treat the Mullen and CDI as two separate snapshots, not copies of the same picture. A low score on one tool deserves a follow-up with the other, plus direct observation. This extra step helps you catch early language delays you might miss if you trust a single number.
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.
After giving either the Mullen or CDI, plan to collect the other within two weeks before setting language goals.
02At a glance
03Original abstract
Infants at elevated likelihood for or later diagnosed with autism typically have smaller vocabularies than their peers, as shown by the Mullen Scales of Early Learning (MSEL) and the MacArthur-Bates Communicative Developmental Inventory (CDI). However, the extent to which MSEL and CDI scores align remains unclear, especially across clinical and non-clinical populations. This study examined whether the concurrent validity of the MSEL and CDI differs based on autism likelihood and diagnosis. Data from 720 14-month-old infants were analysed, grouped by likelihood (elevated vs. typical) and diagnosis at 36 months (diagnosed vs. not diagnosed). Vocabulary scores were compared across both likelihood and diagnostic groups. Moderate correlations were observed between the MSEL and CDI in most groups (rs range = [.34-.58]). One exception was that the expressive scores of elevated likelihood infants on the MSEL and CDI were more closely associated than the expressive scores of typical likelihood infants. Diagnosed infants had lower vocabulary scores than non-diagnosed peers on both the MSEL and CDI. The elevated likelihood group showed lower scores on the MSEL but not the CDI compared to typical likelihood infants. The moderate correlations suggest that the MSEL and CDI assess different aspects of language in infancy. These associations were weaker than previously reported in autistic children. Differences in vocabulary scores across likelihood and diagnostic groups highlight the need for further research to understand the association between these measures.
Journal of autism and developmental disorders, 2025 · doi:10.1111/1473-4192.00024