Modelos de Regressão na Análise de Dados em Ciências Sociais

Imagem d@ Maria Eugénia Ferrão [Ficheiro Local]
Maria Eugénia Ferrão
Código:
17568
Ano:
1
Semestre:
L0
Créditos ECTS:
1
Carga Horária:
OT(17H)/TP(10H)
Área Científica:
Ciências Sociais
Objectivos de Aprendizagem:
This course offers an introduction to statistical regression modeling, focusing on various regression techniques widely used in the social sciences. Key topics include linear regression, binary logistic regression for dichotomous outcomes, and multinomial and ordinal logistic regression for categorical outcomes. Illustrative examples are drawn from a range of social science disciplines. The course also features computer lab sessions utilizing the R software for data analysis, with no prior experience in R required.
The main objectives are as follows: (i) Understand the fundamentals of statistical regression modeling relevant to the social sciences. (ii) Gain knowledge of key regression techniques, including: Linear regression; Binary logistic regression for dichotomous outcomes; Multinomial and ordinal logistic regression for categorical outcomes. (iii)Apply regression models to real-world data from various social science disciplines. (iv) Develop practical skills in using R software.
Conteúdos programáticos:
Fundamentals of R
Introduction to R and RStudio;
Basic structure of the R language;
Vectors, matrices, lists, and data frames;
Data import and organization.

Data Manipulation and Multiple Linear Regression
Data manipulation and preprocessing;
Descriptive statistics;
Multiple linear regression;
Interpretation of coefficients;
Introduction to Generalized Linear Models (GLMs).

Logit, Probit, and Ordinal Models
Theoretical foundations of logit and probit models;
Practical applications in R;
Interpretation of odds ratios;
Theoretical and practical introduction to ordinal models.

Multinomial Models and Model Evaluation
Multinomial logistic regression;
Interpretation of marginal effects;
Goodness-of-fit tests and model quality assessment;
Exploration of R's built-in datasets.
Metodologias de Ensino e Critérios de Avaliação:
Based on a case study presented and defended in a seminar setting.
Bibliografia principal:
Wickham, H., Çetinkaya-Rundel, M., & Grolemund, G. (2023). R for Data Science (2nd ed.). O'Reilly Media.
Língua:
Português