Quantitative Analysis

Design Your Observational Study with the Joint Variable Importance Plot

March 12, 2024
by Lauren Liao. When evaluating causal inference in observational studies, there often is a natural imbalance in the data. Luckily, variables are often measured alongside that can be helpful for adjustment. However, deciding which variables should be prioritized for adjustment is not trivial – since not all variables are equally important to the intervention or the outcome. I recommend using the joint variable importance plot during the observational study design phase to visualize which variables should be prioritized. This post provides a gentle guide on how to do so and why it is important.

A Basic Introduction to Hierarchical Linear Modeling

March 4, 2024
by Mingfeng Xue. Hierarchical Linear Modeling (HLM) is an extension of linear models, which offers an approach to analyzing data structures with nested levels. This blog elucidates HLM's significance over traditional linear regression models, particularly in handling clustered data and multilevel predictors. Illustrated with an example from educational research, the blog demonstrates model implementation and interpretation steps. It showcases how HLM accommodates both independent variables from different levels and hierarchical structure data, providing insights into their impacts on the outcome variable. Recommended resources further aid readers in mastering HLM techniques.

From Asking Causal Questions to Making Causal Inference

December 5, 2023
by Lauren Liao. What is causality and how do we ask causal questions? It may seem like a difficult and foreign concept, but fear not, I will guide you through the basic concepts in this blog post. We will start from how to ask causal questions then more formally address how to answer these questions. You may find causality more approachable than you think. It follows the same ideas as presented by the scientific method of rigorously testing how interventions produce different outcomes in a controlled environment.

Searching for Other Solar Systems

November 21, 2023
by Emma Turtelboom. Over the last three decades, we have discovered over 5000 exoplanets, which are planets outside of our Solar System. With these observations, we can try to answer many questions we have about the universe. For example, how unique is the Solar System? How do planets form? Is there life elsewhere in the Milky Way? We can query the NASA Exoplanet Archive to compare multi-planet systems to the Solar System. Through this, we can compare how similar (or dissimilar!) the systems are.

Exploratory Data Analysis in Social Science Research

November 14, 2023
by Kamya Yadav. Causal inference has become the dominant endeavor for many political scientists, often at the expense of good research questions and theory building. Returning to descriptive inference – the process of describing the world as it exists – can help formulate research questions worth asking and theory that is grounded in reality. Exploratory data analysis is one method of conducting descriptive inference. It can help social science researchers find empirical patterns and puzzles that motivate their research questions, test correlations between variables, and engage with the existing literature on a topic. In this blog post, I walk through results from exploratory data analysis I conducted for my dissertation project on political ambition of women.

Using Forest Plots to Report Regression Estimates: A Useful Data Visualization Technique

October 17, 2023
by Sharon Green. Regression models help us understand relationships between two or more variables. In many cases, results are summarized in tables that present coefficients, standard errors, and p-values. Reading these can be a slog. Figures such as forest plots can help us communicate results more effectively and may lead to a better understanding of the data. This blog post is a tutorial on two different approaches to creating high-quality and reproducible forest plots, one using ggplot2 and one using the forestplot package.

FSRDC 2023 Annual Meeting and Research Conference

October 2, 2023
by Renee Starowicz. Renee Starowicz, Co-Executive Director of the Berkeley Federal Statistical Research Data Center, provides an overview of the takeaways from the 2023 Annual Federal Statistical Research Data Center Business Meeting and Annual Conference. She provides a brief overview of the Berkeley FSRDC. Then, she describes the priorities for collaboration across national directors to improve outreach to diverse researchers and transparency. Additionally, she points out the other key topics of conversation at this year’s meeting.

Introduction to Field Experiments and Randomized Controlled Trials

July 24, 2023
by Leena Bhai. This blog post provides an introduction to field experimentation and its significance in understanding cause and effect. It explains how randomized experiments represent an unbiased method for determining what works. It delves into essential features of experiments such as intervention, excludability, and non-interference. It then works through a fictional example of a randomized controlled trial of the efficacy of an experimental drug Covi-Mapp.