Zhaowen Guo - Statistics

Balanced Cluster Randomization: Strategies for Allocating Treatment and Control Groups

experimental design

Experimental designs sometimes require randomization at the cluster level, where groups of individuals—rather than individuals themselves—are assigned to treatment or…

Weights in Statistics: What Do People Often Get Wrong?

survey methodology

Survey weighting is a mess. — Andrew Gelman

Hype or Harm: Examining Media Coverage of AI

machine learning

natural language processing

Artificial Intelligence (AI) and Large Language Models (LLMs) have become buzzwords, sparking significant media attention. I found myself intrigued by how AI is portrayed in…

Scraping Dynamic Websites with R: An Example from Media Bias Data

natural language processing

In my recent analysis of how news headlines on artificial intelligence (AI) vary across media outlets with different ideological leanings, I stumbled upon an interesting data…

The Power of Singular Vector Decomposition: A Beginner’s Guide

machine learning

natural language processing

SVD is not nearly as famous as it should be. — Gilbert Strang

Revisiting the Monty Hall Problem: The Power of Simulations

Named after the…

Explaining the Bias-Variance Tradeoff

machine learning

In order to assess the model performance on a given dataset, we need to measure how closely its predictions align with the actual data. In regression settings, we often use mean squared error…

Predicting Household Groupings: A Longitudinal Approach

Poverty is a complex issue that is typically measured at the household level, taking into account the resources that are shared among multiple individuals. However, many…

Unveilling Model Equivalence: Linking the Rasch Model with the Hierarchical Linear Model

When learning more about statistical modeling, I find it helpful to consider the relationships and connections between different models. This allows me to gain a better…