OBJECTIVES: to assess the validity of the Movement Assessment of Infants (MAI) test, comprised of 65 items, grouped into four sections, used to detect cerebral palsy. METHODS: the sample included 73 preterm babies, assessed between four and eight months of corrected age. Individual assessments were made for the two ages, using Rasch's statistical model. RESULTS: MAI is an easy scale for the level of ability of the babies assessed, including both age groups, four and eight months old. Some of the items do not fit the one-dimensional expectations of the statistical model, compromising test construction validity. CONCLUSIONS: differences in the difficulty level of the items at four and eight months old reaffirm the need of different performance profiles for each age. Notwithstanding the restrictions, the test has positive aspects and review suggestions are offered aiming at validity improvement.
Validity of tests; Musculoskeletal development; Cerebral palsy