Category Archives: Computational Social Science

The Synthetic Control Method (SCM) is a statistical approach for estimating the causal effect of a treatment in comparative case studies. It is particularly suited for a case where there is one treated unit (e.g., village, state, country) and multiple untreated units observed across time. Therefore, it is applicable to aggregated time-series data. The methodContinue reading “Quasi-Experimental Design: Synthetic Control Method”

Natural Experiment Before discussing natural experiment, it makes sense to explain randomized assignment. By this, I mean that we randomly assign eligible units (or observations) to either a treatment or a control group. Due to the random assignment, there is a 50-50 chance for each eligible observation to be selected, which leads to an internallyContinue reading “Quasi-Experimental Design: Natural Experiment and Instrumental Variable”

Regression discontinuity design (RDD) is a quasi-experimental method for causal inference. RDD has become a prominent method for causal inference since the 2000s. One can account for both observed and unobserved heterogeneity. We can leverage our knowledge of treatment assignment to estimate the effect of a treatment, such as policy, an intervention, on an outcome.Continue reading “Quasi-Experimental Design: Regression Discontinuity Design”

I have completed part I (chapter 1-17) and part II (chapter 18 – 46) of Practical Tableau by Ryan Sleeper.  I think Alexander’s book and Ryan’s book complement each other. Alexander succeeded in focusing on the fundamentals of Tableau and kept the page count modest.  It definitely gave me a good start with Tableau.  Continue reading “Tableau for Data Visualisation: Part 2”

Univariate Timeseries Modelling (UTM) is simply the modelling and prediction of future economic variables using information from own past values and present/previous residual values. In other words, the values of y-variable in past periods (along with the current residual) influences the value of y-variable. We neither have explanatory x-variables nor other explained y-variables; we hardlyContinue reading “Econometrics: Univariate Timeseries Modelling and Forecasting”

Although I love to use R for customization and visualisation, certain graphs are much more easy and intuitive in Tableau because most of its interactions are achieved simply by dragging and dropping items onto the canvas. I recently ordered two books, which I am using to learn Tableau. Visual Analytics with Tableau by Alexander LothContinue reading “Tableau for Data Visualisation”

The term matching is a procedure where we try to find for a sample observation other observations in the sample that have similar observable features. We often refer to the selected observations as matches, and after a repetition of this procedure for all or part of these observations, the subsample of observations that come out of this isContinue reading “Quasi-Experimental Design: Genetic Matching”

I am not a data scientist. But I have always wondered what the term data science meant when I began learning statistics with R. According to a youtube video by Professor Dr. Robert Curtis: Data science = Programming + Statistics + Mathematics However, I will add Communication. I also thought about which resources you canContinue reading “Data Science with R”

What is text mining, and where does it fit? Text mining, also referred to as text analytics, involves the extraction of insightful information from text, which can help in making well-informed decisions. The method is used in academia (e.g. to analytically understand transcripts that are qualitatively collected), data science (e.g. to produce relevant inputs forContinue reading “Text Mining: Using Text as Data”