Analiza logarytmiczno-liniowa w badaniu przyczyn umieralności w krajach UE
Log-linear analysis in the study of mortality in EU

Summary/Abstract: Log-linear models are a widely used tool for modeling qualitative data in contingency table. Hierarchical log-linear models include lower order terms implied by any higher order ones. The fit of log-linear model can be assessed with the Pearson or likelihood- ratio chi-square, as well as information criteria: AIC [Akaike 1973], BIC [Raftery 1986] and Aitkin`s method. Identifying the simplest model with the fewest parameters is the main goal of the analysis. Log-linear analysis is available in R software with the use of loglm function in library MASS. The empirical use of log-linear analysis is based on European Union countries mortality dataset.

• Details
• Contents