Research Seminar "Quantitative Methods in Political Research"
Камалов Эмиль Альфредович
- To acquaint students with statistical methods and terminology
- To teach students how to implement statistical methods using R programming language.
- reads (and understands) most academic PS articles
- speaks the language of data fluently
- designs a quantitative political study
- chooses statistical methods appropriate to his substantive research problem
- uses R programming language for statistical computations
- Design types, data types, and data summarization1.1. Prospective cohort studies, case/control studies, cross-sectional studies 1.2. Quantitative research design 1.3. Continuous and nominal-categorical data, ordered data, summary statistics, visualization 1.4. Try R, R as calculator, basic R functions, creation of the data frame
- Basic Statistical Concepts2.1. Population vs. Sample 2.2. Random Sample 2.3. Probability, probability distributions 2.4. Work with R functions: pnorm, tnorm, density, etc.
- Exploratory Data Analysis and Visualization3.1. Work with R: set working directory, read your data, data types in R 3.2. Data transformation, missing values, outliers 3.3. Visualization: histograms, barplots, boxplots
- Inference and Hypothesis Testing4.1. Normal distribution, central limit theorem 4.2. Null hypothesis, alternative hypothesis 4.3. Type I error, type II error, P-value 4.4. Confidence intervals 4.5. Causal inference
- Simple Regression Methods5.1. Linear regression assumptions 5.2. OLS regression, regression coefficients, significance level alpha 5.3. OLS regression example in R: syntax and interpretation, visualization of the result
- Confounding and Effect Modification (Interaction)6.1. Define and identify confounding 6.2. Crude and adjusted measures 6.3. An example of effect modification 6.4. Define and identify interaction effect 6.5. An example of interaction effect
- Multiple Regression Methods7.1. Diagnostic plots 7.2. Comparing model 7.3. Variable selection 7.4. Tools for summarizing and visualizing regression models
- Generalized Linear Models 18.1. Logit regression 8.2. Poisson regression 8.3. Fitting and visualizing Logit and Poisson regressions using R
- Generalized Linear Models 29.1. Ordered regressions 9.2. Multinomial regression 9.3. Fitting and visualizing Ordered and Multinomial regressions using R
- Homework (1-5)The most important aspects of assignments that affect grades are following: a) correctness of answers to questions given in an assignment, b) ability to write R code correctly (if necessary), c) appropriate use of statistical language, d) correctness of results’ interpretations. If all these criterions are met, you can expect an excellent grade (8-10 on 0-10 scale). Late assignments will be graded down by 1 point for each day of delay (but no more than 3 points in total). Plagiarism is prohibited.
- In-class Participation
- Midterm paperThe most important aspects of the paper to be graded are: 1) logical reasoning, 2) correctness and efficiency of R code written, 3) accuracy of statistical methods and models used, 4) correctness and creativity of results’ interpretation, 5) fluency and accuracy of statistical terminology used.
- Final paperThe most important aspects of the paper to be graded are: 1) logical reasoning, 2) correctness and efficiency of R code written, 3) accuracy of statistical methods and models used, 4) correctness and creativity of results’ interpretation, 5) fluency and accuracy of statistical terminology used.
- Interim assessment (4 module)0.3 * Final paper + 0.25 * Homework (1-5) + 0.25 * In-class Participation + 0.2 * Midterm paper
- Wilcox, R. R. (2016). Understanding and Applying Basic Statistical Methods Using R. Hoboken, New Jersey: Wiley. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=1237377
- Val Todd. (2017). Field and Iles (2016) An Adventure in Statistics: The Reality Enigma. PRISM, 1(1), 195–199. https://doi.org/10.24377/LJMU.prism.vol1iss1article304