Uso de modelos estadísticos para la estimación de tasas de desocupación en dominios pequeños

Abstract

The parameter estimates in some domains or areas could have some problems when the simple size corresponding to those domains is reduced as the “direct estimates” could be unstable. The random effects models are a methodological alternative to improve these estimates. In this article the random effects models are used to estimate the rate of unemployment in small domains with the data provided by the Permanent Household Survey (EPH) corresponding to the Gran Rosario region. A simulation study is performed to extend the results to other scenarios of analysis. The results founded demonstrate the convenience of the use of the estimates based on models when the sample size of a group and their variability are small.