Perhaps no other subject in applied statistics and machine learning has caused people as much trouble as the humble logit model. Most people who learn either subject start off with some form of linear regression, and having grasped it, are then told... [Read more...]

Robustness check: time of completion Survey Design Sensitivity Estimates Alternative Estimation: RRreg Alternative Estimator: rr BRMS Model Adjustment with MRP Gender Age Region Simulation Comparison: Univariate Vs. Regression References This ... [Read more...]

This post focuses on one of the more curious models in contemporary statistics, a specification for proportions that is either called fractional logit or quasi-Binomial. While in most cases in statistics, the evaluation of a model necessarily invol... [Read more...]

Twitter (or what’s left of it) was recently ablaze with a discussion of two smart working papers, one by Jiafeng Chen and Jonathan Roth and the other by John Mullahy and Edward Norton. In different ways, they argue against the use of non-linear t... [Read more...]

I am writing this post in response to questions about estimating turnout for Tunisia’s constitutional referendum today. Turnout is an important aspect to this referendum because high turnout would signal higher legitimacy for President Kais Saied’s... [Read more...]

Introduction Limited dependent variables, or continuous variables with lower and upper bounds, are quite common in the social sciences but do not fit easily with existing statistical models. In this Rmarkdown document, I show why these issues are ... [Read more...]

NB: An earlier version of this post critiqued Victor Chernozhukov’s approach to directed a-cyclic graphs and fixed effects, but made some critical errors in interpreting his approach. These errors were entirely mine, and I apologize to Victor for d... [Read more...]

There has been plenty of discussion about declining fertility rates and patterns of marriage among people in the United States following the news that the US birth rate declined to its lowest since the Great Depression. There are a lot of debates a... [Read more...]

This tutorial gives an overview of the COVID-19 policy indexes just released by the CoronaNet project of which I am a part and the Oxford Government Response Tracker. See the associated paper here. This post shows you how to download the estimate... [Read more...]

The COVID-19 pandemic has led to a wealth of research studies examining various aspects of the pandemic. In this post, I will discuss some of the available datasets for doing aggregate level analysis, i.e., comparing countries, provinces, or counti... [Read more...]

Recently I had a spirited conversation with Victor Chernozhukov, a leading econometrician whose work spans both traditional subjects like panel data modeling and newer application of machine learning models to causal inference. While I admire his w... [Read more...]

While perusing the news, I read eagerly about the CDC’s recent study examining associations between county-level mask mandates and COVID-19 growth rates. This study has already been the subject of angry retorts from the restaurant industry due to the CDC’s claim that restaurant closures reduced COVID-19 spread. Looking ... [Read more...]

In this blog post, I use Gelman and Loken’s garden of forking paths analysis to construct a simulation showing why skepticism of AstraZeneca’s vaccine results is warranted at this early stage. This simulation uses the numbers from their press relea... [Read more...]

I start with Eric Novik’s excellent blog post on how to calculate the relevant statistics for the vaccine, i.e. vaccine efficacy (VE). This is defined as: \[ VE = 1 - \frac{p_t}{p_c} \] Where \(p_t\) is the proportion of cases in the treatment (vac... [Read more...]

My wife and I have been faced with a decision in our pregnancies that has always caused me some consternation: should we take the MaterniT 21 test to see if our baby might have Down’s syndrome (trisomy 21) or other genetic abnormalities? This tes... [Read more...]

We’ve all been in that seminar where the author puts up a slide containing regression coefficients, and buried in the bottom line of the table we can see little Ys and Ns indicating the kind of panel data model employed. Quite often, these brief descriptions of models are taken ... [Read more...]

For an up to date version of this model, please see our paper at https://osf.io/preprints/socarxiv/jp4wk/. [Read more...]