Generalized Linear Models - DEs/UEM

Final work developed for the course of Generalized Linear Models in the first semester of 2017 taught by professor Rosangela Getirana Santana at DEs/UEM.

Regression models: Applications in counting and proportions data

This work presents the analyze of two datasets in which the response variables are discrete from a given count, and restricted to the interval (0, 1).

• The first dataset is from Long (1990) and the response variable is the number of publications of Ph.D. students in the last three years. The main goal of this analysis is to determine which factors affect the produtivity of the Ph.D students.

• The second dataset refer to the proportion of valid votes that the former president Dilma received in the second round of the 2010 election, considering the cities of Parana state. In this analysis the goal is to assess which were the main factors that influence the proportion of votes obtained by Dilma.

---

Report

Presentation

R scripts

SAS scripts

---