Tesis doctorales: TESEO

Ayuda Salir

Ficha de Tesis

Imprimir

Ministerio de Universidades - Gobierno de España

Título: SMALL AREA ESTIMATION. AN APPLICATION TO THE ESTIMATION OF THE LABOUR MARKET VARIABLES IN GALICIA COUNITES
Autor: López Vizcaíno, Mª Esther
Universidad: Universidad de A Coruña
Departamento: Matemáticas
Fecha de Lectura: 25/04/2014
Programa de doctorado: Estadística e inestigación operativa
Dirección:
- Lombardía Cortiña, Mª José (DIRECTOR)
- Vaamonde Liste, Antonio (CODIRECTOR)
- MORALES GONZALEZ, DOMINGO (CODIRECTOR)
Tribunal:
- GONZÁLEZ MANTEIGA, WENCESLAO (PRESIDENTE)
- Vilar Fernández, Juan Manuel (SECRETARIO)
- RUEDA SABATER, CRISTINA (VOCAL)
- MOLINA PERALTA, ISABEL (VOCAL)
- UGARTE MARTINEZ, MARÍA DOLORES (VOCAL)
Descriptores:
- TEORIA Y TECNICAS DE MUESTREO
- ANALISIS DE DATOS
- TECNICAS DE INFERENCIA ESTADISTICA
Localización: REPOSITORIO BIBLIOTECA UNIVERSIDAD DE A CORUÑA

http://hdl.handle.net/2183/12166

Marcador: https://www.educacion.gob.es/teseo/mostrarRef.do?ref=1088895Añadir a favoritos.
Resumen:
Unemployment is currently one of the most important problems of our society. Global measures to fight against unemployment are usually not satisfactory to carry out effective policies at the local level, hence the need of tools to provide accurate labor market information at local level. In Spain the estimation of labor market indicators is made by means of the Labour Force Survey that is designed to obtain information at the provincial level. Sample sizes below the provincial level of disaggregation are low and direct estimators in the counties or municipalities often have low precision. Therefore, the aim of this thesis is to estimate labor indicators, such as employed, unemployed and unemployment rates, in Galician counties. We use small area estimation techniques under area level multinomial mixed models. We will first use estimators based on the area level multinomial mixed model introduced by Molina et al. (2007). These authors considered a model with a common random effect for the two multinomial categories (employed and unemployed people). In the real data problem this may not be appropriate because of the very different characteristics of these two populations. This is the reason why we introduce in the model two random effects, one for each of the multinomial categories. In addition, the availability of surveys for different periods of time produces a significant increase of the domain samples and allows us to introduce in the model independent and correlated time random effects. The estimate of the accuracy of the estimators of small areas is a fundamental issue because these estimators are often biased. In this work we use different methods for estimating the mean squared error, by using analytical expressions and using bootstrap techniques

Imprimir

Modificar Consulta Nueva Consulta

Versión 4.4