Abstract
Life expectancy among the elderly has been increasing for decades and it has drawn attention to the health conditions of those added years. Functional Capacity (FC) has been used as a basic indicator to appraise health conditions and elderly well-being.
To conceptually characterize FC, there is a standardized framework such as The International Classification of Functioning, Disability and Health (ICF). The diversity of proposals to operationalize the concept covers from the way the question is constructed to the scales used to evaluate FC, making it difficult comparison among studies.
An indicator for FC is constructed using Item Response Theory (IRT). Data is the health supplement of the Brazilian National Household Survey (PNAD) in 1998, 2003 and 2008. Since the seven questions under scrutiny presented several level answers, a Generalized Partial Credit Model for ordered polytomous data is adjusted. Answers included four levels of difficulty. A GLM model was then fitted to the indicator using socio-economic explanatory variables.
This paper shows that it is possible to construct a FC scale using IRT. Comparison between two populations is possible if at least a common question exists for the two surveys. One can classify tasks according to some power of discrimination between levels of FC.
To conceptually characterize FC, there is a standardized framework such as The International Classification of Functioning, Disability and Health (ICF). The diversity of proposals to operationalize the concept covers from the way the question is constructed to the scales used to evaluate FC, making it difficult comparison among studies.
An indicator for FC is constructed using Item Response Theory (IRT). Data is the health supplement of the Brazilian National Household Survey (PNAD) in 1998, 2003 and 2008. Since the seven questions under scrutiny presented several level answers, a Generalized Partial Credit Model for ordered polytomous data is adjusted. Answers included four levels of difficulty. A GLM model was then fitted to the indicator using socio-economic explanatory variables.
This paper shows that it is possible to construct a FC scale using IRT. Comparison between two populations is possible if at least a common question exists for the two surveys. One can classify tasks according to some power of discrimination between levels of FC.
confirm funding
Event ID
17
Paper presenter
47 038
Type of Submissions
Regular session presentation, if not selected I agree to present my paper as a poster
Language of Presentation
English
First Choice History
Initial First Choice
Weight in Programme
3
Status in Programme
1