Statistical Methods in Social Sciences

Content

Objective

This course aims to help students conducting research in the social sciences and international relations acquire the statistical mindset necessary to manage the uncertainty they encounter. The course requires no prior programming knowledge and begins by building skills in data manipulation, cleaning, transformation, and visualization with modern graphical techniques in R; the logic of reproducible analysis via R Markdown/Quarto is introduced early on.

In subsequent weeks, simple and multiple linear regression models are constructed and interpreted in line with research questions, and group differences and conditional relationships are analyzed through dummy variables and interaction terms. At the end of the semester, the basic logic of causal inference and the difference-in-differences approach are introduced; the panel data structure and how unobserved unit heterogeneity is addressed through fixed effects, first differences, and random effects are demonstrated. Throughout the course, students are expected to critically interpret statistical inference tools such as standard errors, confidence intervals, p-values, and effect sizes, and to transform their results into valid insights and empirical evidence.

In this course, students are not expected to memorize formulas; rather, they are expected to select data appropriate to the research question, analyze it in the R environment, critically interpret the results, carry out a joint analysis process within a group, and report this process in a reproducible manner. Students may use AI tools in coding, debugging, and interpretation, but they are obliged to explain and defend every analysis step they use.

Course Content

The course begins with setting up a reproducible data analysis process in the R/RStudio and Quarto/R Markdown environment. It then covers data management, descriptive analysis, and visualization with ggplot2. Simple and multiple linear regression and statistical inference are then introduced. At the end of the semester, the difference-in-differences approach and panel data analysis are introduced.

Course Learning Outcomes

Upon successful completion of this course, the student will be able to:
1. Set up a reproducible data analysis process in the R/RStudio and R Markdown/Quarto environment; carry out data loading, cleaning, transformation, and reporting steps in an integrated manner.
2. Organize a real social science dataset, explore it with descriptive statistics and ggplot2 graphics, and interpret the basic patterns in the data contextually.
3. Construct simple and multiple linear regression models, interpret coefficients, explain the role of control variables, and evaluate model outputs.
4. Perform statistical inference and apply hypothesis tests.
5. Analyze group differences and conditional relationships within a regression framework using dummy variables and interaction terms; support the results with graphs and interpret them.
6. Explain the basic difference-in-differences approach; discuss treatment and control groups, beforeafter comparison, and the parallel trends assumption.
7. Explain the panel data structure; discuss unobserved unit heterogeneity and, on an appropriate dataset, apply one of the fixed effects, first differences, or random effects approaches in R and report the findings in scientific language.
8. Define analysis responsibilities in the contribution statement in paired group work, contribute to the joint report

Teaching and Learning Methods

• Computer-assisted applied instruction: live data analysis, data cleaning, visualization, and reporting with R/RStudio.
• Conceptual discussion and intuitive learning: classroom dialogue centered on research questions, method selection, and interpretation rather than formula memorization.
• Case analysis on real-world social science datasets: V-Dem, World Bank, UCDP, Gapminder, World Values Survey, European Social Survey, etc.
• Short, application-oriented learning activities with instant feedback; students may use AI during the analysis process, but must be able to explain and defend every step used.
• Paired group work: students share responsibilities in data preparation, analysis, reporting, and poster presentation; critically evaluate each other’s code, tables, figures, and interpretations.

References

1. Wickham, H., Çetinkaya-Rundel, M., & Grolemund, G. (2023). R for Data Science (2nd ed.). Open-access foundational resource for data management, cleaning, transformation, and visualization with R.
https://r4ds.hadley.nz/ (English) and https://tr.r4ds.hadley.nz/ (Turkish)
2. Wooldridge, J. M. (2020). Introductory Econometrics: A Modern Approach (7th ed.). Cengage Learning. For this course, reading the first 4 chapters and chapters 6, 7, 13, and 14 is sufficient. For the datasets and R code used in the book, see
https://justinmshea.github.io/wooldridge/
3. Some open datasets and R packages to be used throughout the course:
• V-Dem Institute: https://www.v-dem.net/data/, R package: vdemdata
• World Bank Open Data: https://data.worldbank.org/, R packages: WDI, worldbank.
• UCDP/PRIO Armed Conflict Dataset: https://www.prio.org/data/4
• Gapminder: R package: gapminder
• World Values Survey: https://www.worldvaluessurvey.org/
• European Social Survey: https://www.europeansocialsurvey.org/, R package: essurvey
• For other data easily downloadable and

Weekly Contents

Assessment System

Relation of Proficiency

ECTS