This article aims to evaluate the items used to measure functional capacity in the 2010 Health Survey in Greater Metropolitan Belo Horizonte, Minas Gerais State, Brazil, using Item Response Theory. We analyzed the scale’s dimensionality, the items’ position and discriminatory power, and the precision of the functional capacity estimate. The study was based on a sample of 2,174 individuals aged 60 to 99 years, 61% of whom were women, with a median of 4 years of schooling. The instrument with 21 items (4 response options) showed a Cronbach’s alpha coefficient of 0.98. Factor analysis identified one factor that explained 92% of the variability between the items. The results indicated: (i) redundancy in items; (ii) precise estimation of functional capacity only for elderly below the median on the scale; and (iii) inability of the elderly to differentiate between response options. The analysis suggests the need to discuss items included in the instrument in order to cover a wider range of the scale and seek more appropriate response options.
Health of the Elderly; Evaluation of Research Programs and Tools; Multivariate Analysis
