Servicios Personalizados
Revista
Articulo
Indicadores
- Citado por SciELO
- Accesos
Links relacionados
- Citado por Google
- Similares en SciELO
- Similares en Google
Compartir
Revista Facultad Nacional de Salud Pública
versión impresa ISSN 0120-386Xversión On-line ISSN 2256-3334
Resumen
VALENCIA C, Marisol y SALAZAR U, Juan C. Impact evaluation of welfare actions over a community in Colombia, using a data correlated model. Rev. Fac. Nac. Salud Pública [online]. 2010, vol.28, n.1, pp.64-72. ISSN 0120-386X.
The impact evaluation of a communitarian program sometimes is based on statistical techniques such as linear regression models analysis when the goal is usually to quantify its effects on both social and welfare characteristics of a community. This involves the study of mental health and social variables which effects could not be easily quantified due to the presence of correlation structures among the observations within a particular subject. For this reason it is advisable to use linear mixed models to this kind of study. OBJECTIVE: To calculate the impact of the effects produced by actions of the Instituto de Deportes y Recreación de Medellín (INDER) on the participant population. METHODOLOGY: from a sample collected by INDER back in 2007, about participants and no participants of its programs, generalized linear models were estimated to explain the behavior of both social and psychological variables. Then, by using a logistic regression model a matching procedure was performed to identify the subjects and their repeated measures that will serve as inputs to measure the impact of the intervention of the welfare activities on the community by means of a linear mixed model. RESULTS: The linear mixed estimation process identified important interaction variables that favor the intervention welfare actions. CONCLUSIONS: from the participation in the INDER activities, an improvement of variables related with social capital was found.
Palabras clave : Linear mixed models; mental health; scores for impact evaluation.