Panel Data: Analysis and Applications for the Social Sciences
- The course aims to provide students with the theoretical background and practical skills in conducting panel data analysis. Specifically, the learning objectives are as follows: to enable students to choose appropriate models for panel data analysis to develop data manipulation and visualization skills to enable students to implement linear panel models in RStudio
- By the end of the course students are expected to apply fixed- and random- effects models to analyze panel data, to interpret the results, to have data visualization skills and skills in implementing the afore-mentioned methods by using RStudio in the context of panel data analysis. Students will learn the advantages and limitations of different approaches to panel data analysis. This knowledge will help students choose a set of appropriate statistical tools to test their research hypotheses.
- Introduction. Linear regression analysisTypes of data structures. Multiple linear regression models with their applications to crosssectional data. Assumptions. Model specification. Interpretation of regression analysis results. Model diagnostics.
- Data manipulation. Supplementary tools for panel data analysisPanel VS Time-series cross-section (TSCS) VS Time-series data. Exploratory data analysis and visualization of panel data. Within- and between-group variation. Reshaping data. Merging data. Students are required to listen to the following lectures online (Week 3, Week 4, “Getting and Cleaning Data”. Available at: https://www.coursera.org/learn/data-cleaning) before the given practical session.
- Interaction terms in regression analysisModeration VS Mediation. Conditional hypotheses with examples from social science research. Multiple linear regression models with interaction terms. Model specification. Interpretation of interaction effects. Interaction between binary predictors. Interaction between binary and continuous predictors. Marginal effects. Visualization of interaction effects.
- Fixed-effects modelsFixed-effects model VS pooled model. Least-squares dummy-variable models. Within-group transformation. The technique underlying the estimation of coefficients in fixed-effects models. Aggregation bias. Model diagnostics.
- Random-effects models VS Fixed-effects modelsRandom-effects models: assumptions, model estimation, generalized least-squares method and feasible generalized least-squares method. Hausman test and its limitations.
- Interim assessment (3 module)0.3 * 3 home assignments + 0.25 * Quantitative research essay + 0.3 * Quizzes + 0.15 * Seminar activity
- Charles N. Halaby. (2003). Running Head: Panel Models Panel Models in Sociological Research: Theory into Practice. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsbas&AN=edsbas.72A0015F
- Анализ панельных данных и данных о длительности состояний : учеб. пособие, Ратникова Т. А., Фурманов К. К., 2014
- Эконометрика: Начальный курс : учебник для вузов, Магнус Я. Р., Катышев П. К., 2001