|
|
Applied multivariate statistical analysis (second term 2017/18)
Structure of the course
Course aims and objectives
The objective of the course is to familiarize students with the underlying assumptions of the main statistical techniques for data analysis used in political science so that they will be able to evaluate and undertake quantitative research by their own. A great emphasis will be placed on the formulation of hypotheses and on the use of data to test hypotheses. The first half of the course is devoted to test non-linear models with Ordinary Least Squares (OLS) and to to discuss about fixed effect models. In the second half of the course we will introduce some more advanced techniques for quantitative analysis (random models and probit/logit). Lectures are coordinated with computer lab instruction in data analysis. Students will also learn how to use the statistical software STATA to organize and analyze data.
Course prerequisites
The students must be familiar with the basic concepts of descriptive and inferential statistics (levels of measurement, probability, hypothesis testing, confidence interval).
Required reading
Stock J.H and M. W. Watson (2003), Introduction to Econometrics, Boston: Addison Wesley
Lectures
Ten lectures will take place (five for the first half; five for the second half)
Examination
Course grades will be based on two written exam (the first after the first half of the course; the second at the end of the course).
The objective of the course is to familiarize students with the underlying assumptions of the main statistical techniques for data analysis used in political science so that they will be able to evaluate and undertake quantitative research by their own. A great emphasis will be placed on the formulation of hypotheses and on the use of data to test hypotheses. The first half of the course is devoted to test non-linear models with Ordinary Least Squares (OLS) and to to discuss about fixed effect models. In the second half of the course we will introduce some more advanced techniques for quantitative analysis (random models and probit/logit). Lectures are coordinated with computer lab instruction in data analysis. Students will also learn how to use the statistical software STATA to organize and analyze data.
Course prerequisites
The students must be familiar with the basic concepts of descriptive and inferential statistics (levels of measurement, probability, hypothesis testing, confidence interval).
Required reading
Stock J.H and M. W. Watson (2003), Introduction to Econometrics, Boston: Addison Wesley
Lectures
Ten lectures will take place (five for the first half; five for the second half)
Examination
Course grades will be based on two written exam (the first after the first half of the course; the second at the end of the course).
First theme: Interaction Models (first half of the course)
Notes on Interaction Models with OLS
Dataset 1 (NES 2004)
Replication do-file
Second theme: Margins Command (first half of the course)
Notes on Margins
Dataset 1 (Lijphart)
Dataset 2 (NES 2004)
Dataset 3 (Caschool)
Dataset 4 (SWD macro)
Replication do-file
Third theme: Fixed Effects Linear Models (first part)
Notes on issue of Independence (first part)
Dataset 1 (Consumption)
Dataset 2 (SWD micro)
Dataset 3 (Smoking)
Dataset 4 (Happiness)
Replication do-file
First half exam (script)
Fourth theme: Random Linear Models (second part)
Notes on issue of Independence (second part)
Dataset 1 (Consumption)
Dataset 2 (SWD micro)
Dataset 3 (Smoking)
Dataset 4 (Happiness)
Replication do-file
Fifth theme: Regression with a Binary Dependent Variable (second half of the course)
Notes on Probit and Logit
Dataset 1 (School)
Dataset 2 (NES 2004)
Replication do-file
Sixth theme: Regression with a Binary Dependent Variable. Diagnostic (second half of the course)
Notes on Probit and Logit: diagnostic
Dataset 1 (School)
Dataset 2 (NES 2004)
Dataset 3 (Union)
Replication do-file
Second half exam: first dataset; second dataset (script)
Appendix:
Critical values for the chi squared distribution
Cumulative Standard Normal Distribution Function
Notes on Interaction Models with OLS
Dataset 1 (NES 2004)
Replication do-file
Second theme: Margins Command (first half of the course)
Notes on Margins
Dataset 1 (Lijphart)
Dataset 2 (NES 2004)
Dataset 3 (Caschool)
Dataset 4 (SWD macro)
Replication do-file
Third theme: Fixed Effects Linear Models (first part)
Notes on issue of Independence (first part)
Dataset 1 (Consumption)
Dataset 2 (SWD micro)
Dataset 3 (Smoking)
Dataset 4 (Happiness)
Replication do-file
First half exam (script)
Fourth theme: Random Linear Models (second part)
Notes on issue of Independence (second part)
Dataset 1 (Consumption)
Dataset 2 (SWD micro)
Dataset 3 (Smoking)
Dataset 4 (Happiness)
Replication do-file
Fifth theme: Regression with a Binary Dependent Variable (second half of the course)
Notes on Probit and Logit
Dataset 1 (School)
Dataset 2 (NES 2004)
Replication do-file
Sixth theme: Regression with a Binary Dependent Variable. Diagnostic (second half of the course)
Notes on Probit and Logit: diagnostic
Dataset 1 (School)
Dataset 2 (NES 2004)
Dataset 3 (Union)
Replication do-file
Second half exam: first dataset; second dataset (script)
Appendix:
Critical values for the chi squared distribution
Cumulative Standard Normal Distribution Function