Vineland-3 Measurement Non-Invariance in Children With and Without Intellectual and Developmental Disabilities.

McClain et al. (2023) asked a simple question. Do Vineland-3 scores mean the same thing for kids with IDD and kids without IDD?

They ran a measurement-invariance test on the Comprehensive Interview Form. The sample mixed youth with intellectual disability, developmental delay, and neurotypical peers.

The test showed configural invariance. In plain words, the same questions load on the same domains for both groups.

But metric invariance looked shaky. A one-point score jump may not reflect the same skill gain in each group. Direct score comparisons are risky.

Thurm et al. (2020) already warned that adaptive scales often miss real change. Their review sets the stage for the new warning.

Halladay (2025) flips the mood. The 2025 review still backs Vineland-3 for ID trials, urging person-ability scores to dodge floor effects.

The two papers seem to clash. One says "watch out, scores may drift," the other says "use it, just score it smarter." The gap is method, not message. Brunson flags a psychometric trap; Alycia offers a scoring workaround.

Nordstrand et al. (2015) adds comfort. Their Bayley-III low-verbal study found no item bias in toddlers. Younger kids with delay kept pace, showing invariance can hold when the tool is built for low language.

Before you compare baseline and exit Vineland-3 scores across kids with and without IDD, check the metric. If the scale units stretch differently, a five-point gain may look big for one child and small for another. Run a quick invariance check in your stats package, or use person-ability values instead of raw scores. Either step keeps your progress graphs honest and your treatment decisions solid.