The single-item self-report ratings most used in sport science to monitor training responses have not been validated: in a COSMIN review, 46.1% of the athlete-reported measures in 310 monitoring studies were single items, and apart from 2 studies of reliability and responsiveness no validity study of a single item was found; multi-item questionnaires were mostly adequate, except for content validity and measurement error

Jeffries and colleagues noted that athlete-reported outcome measures are used widely in research and practice, yet their psychometric properties had not been examined.

In part 1 they identified the measures most often used in sport to monitor training responses: of 9446 articles screened by title and abstract, 310 of 334 full texts were included, and 53.9% of the measures had several items while 46.1% were single items.

In part 2 they screened 1895 articles, selected 71, and rated methodological quality, measurement properties and level of evidence with COSMIN. Most measurement properties of the multi-item questionnaires were adequate, but content validity and measurement error were not. Apart from 2 studies of reliability and responsiveness, no validity study of a single item was found.

The authors conclude that the single items most used in sport science are unvalidated, that unvalidated modified versions of those items are common, that conclusions based on them are "questionable" until proper validation is done, and that clinimetric reference methods should be used to develop and test such measures.