Which Score for What? Operationalizing Standardized Cognitive Test Performance for the Assessment of Change.

The analysis used only two rare genetic conditions, PTEN hamartoma tumour syndrome and tuberous sclerosis complex, and excluded the Phelan McDermid Syndrome cohort because only 24 of 101 participants could take the test. Many participants could not take the Stanford-Binet 5, and about half of those without a test were reported to lack sufficient ability, so the findings apply mainly to people who can complete the test. The work tested score types from one cognitive test, the Stanford-Binet 5, so it does not prove the same choice is best for every developmental test. All score types had some model-check problems, including floor effects, ceiling effects, or unusual positive errors, so the statistical conclusions are not free of limitation. The study did not compare treatments and did not ask patients or caregivers whether a score change mattered clinically. The data were from natural history studies, so the results are about measurement and interpretation, not about therapy outcomes.

The paper lists NIH intramural research support. No conflict-of-interest declaration is included in the supplied text.

Do not read the finding that the change sensitive score was most appropriate as proof that it detects clinically meaningful cognitive change. This paper did not compare treatments or ask patients and caregivers whether changes mattered; it tested how different score types behave statistically.