Linear and Generalised Linear Regression
Volledige vakbeschrijvingThis course is a continuation from the course “Social Science Statistics in European Studies”. In the first part of the course we will cover in greater detail the linear regression model. We will see that this model depends heavily on several assumptions. We will examine these assumptions in detail, considering why they are necessary and what the consequences are if they are violated. We will also cover dichotomous independent variables and multiplicative interaction terms. Further, we will demonstrate the composition of the independent variable effect. This will form the foundation for structural equation modeling. Additionally, we will cover the step-wise regression model and regression diagnostics. The second part of the course deals with the most fundamental regression models for binary, ordinal, nominal and count outcomes. Particular emphasis is placed on the interpretation of results of these nonlinear models. Various methods of interpretation are presented. The binary model is extended to the multinomial logit model. We will also touch the ordinal logit and probit models.
Doelstellingen van dit vak• read, understand, and evaluate the professional literature that uses advanced linear and generalized linear regression models; • show the ability to independently design and execute the analysis of a research project with multiple regression and generalized linear regression; • be able to thoroughly test for the violation of model assumptions and know how to best present the results of these diagnostic tests in a research publication; • have the competence to choose the appropriate statistical technique for a given research question and data, to properly interpret the results of nonlinear models and to know how to best present these results in a research publication.
Aanbevolen literatuurField, Andy (2009) Discovering Statistics Using SPSS. Sage. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences, 3rd Edition. London: Lawrence Erlbaum Associates. Long, J. Scott. 1997. Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, CA: Sage.
25 okt 2021
17 dec 2021
Taal van de opleiding:Engels
- L. Russo
Trefwoorden:Linear and generalized linear regression, model assumptions, model diagnostics and interpretation, binary logistic regression, ordinal logit and probit models, multinomial logistic, count data.