¿Se pueden predecir geográficamente los resultados electorales? Una aplicación del análisis de clusters y outliers espaciales

The results of this study show that using spatial statistics in electoral geography can predict electoral results. The geographical concepts of spatial cluster and spatial outlier are applied, and a local spatial segregation measure used as the predictor variable. The statistical techniques employed...

Description complète

Détails bibliographiques
Auteur principal: Vilalta Perdomo, Carlos J.
Format: Online
Langue:espagnol
Éditeur: El Colegio de México A.C. 2008
Sujets:
Accès en ligne:https://estudiosdemograficosyurbanos.colmex.mx/index.php/edu/article/view/1322
Institution:

Estudios Demográficos y Urbanos

Description
Résumé:The results of this study show that using spatial statistics in electoral geography can predict electoral results. The geographical concepts of spatial cluster and spatial outlier are applied, and a local spatial segregation measure used as the predictor variable. The statistical techniques employed are Moran´s global and local spatial autocorrelation indexes, and linear regression. The analysis shows: 1) that Mexico City contains spatial clusters of electoral support and marginality, 2) spatial outliers of marginality, 3) political parties exclude each other geographically, and 4) electoral results to be significantly dependent on the levels of spatial segregation within the city.