Applied Statistics for Business and Social Science Research

Overview

The objective of this course is to give the students an introduction to advanced econometric methods using Stata. In this course, we will focus on the practical application of the methods used in business research and other social sciences using data relevant for the participants. Being able to write statistical research papers will help PhD candidates in publishing their work in good academic journals. Statistics can also be applied in order to confirm results found in case studies.

We will go through the basics of regression with a focus on diagnostics, before moving on to different ways of modeling when the OLS regression assumptions are breached. This course encompasses both cross-sectional, nested data, and time-series analysis. It differs from other courses in two main ways. First, the topics and examples used will be relevant for students within economics and other fields within social science. Second, the course will focus on applied statistics, thus enabling the candidates to produce a research paper within the framework of the course.

Students will get a repetition about the basics of regression analysis, before moving on to situations where we cannot apply ordinary least squares regression to different types of data. Then will go through different types of dependent variables, including dichotomous variables (logistic regression), variables with more than two categories (multinomial regression), and ordinal variables (ordered logistic regression). The next topic are data that breaches the assumption of independent units, that is, the data are nested. This includes hierarchical or multilevel data and panel data (note that the topics can vary from time to time).

This course will focus on statistical methods relevant for PhD candidates in the social sciences. Research questions within these fields often require the analysis of data where the linear regression assumptions are breached, and where ways of handling nested and time data are required.

Knowledge:

• Demonstrate a knowledge of regression analysis and its limitations;
• To decide which type of regression analysis is appropriate for different types of dependent variables and data structures;
• To understand and use different types of statistical models, including different variants of logistic regression, multilevel modeling, panel data analysis, and time-series-crosssection data;
• Understand, interpret, and present statistical results.

Skills:

• Will be familiar with and able to use the statistical software Stata;
• Prepare data for use;
• Perform an independent statistical analysis;
• Write statistical method sections;
• Write quantitative research papers of good quality.

The course consists of two separate online workshops, which will comprise a mixture of formal presentations and lab work using both example data and the students' own data. The focus of the workshops is to enable the students to be able to use what they learn in the course on their own data relevant for their respective doctoral theses. We will also introduce the students to literature that will aid them in their further work on their statistical models.