The File Drawer


The Strange Selection Bias in LASSO

Let’s suppose you have two variables \(X\) and \(Z\) that are positively correlated. \(X\) causes \(Y\) but \(Z\) does …

Read On →

Can you use IRT estimates in an error-in-variables model? A tentative yes.

One of the laments I often hear in social science is that we don’t take measurement error seriously enough. Fair enough! …

Read On →

Can you use IRT estimates in an error-in-variables model? A tentative yes.

One of the laments I often hear in social science is that we don’t take measurement error seriously enough. Fair enough! …

Read On →

Alcohol and Asteroids

I recently read Emily Oster’s book: Expecting Better: Why the Conventional Pregnancy Wisdom is Wrong—and What You Really …

Read On →

Do cluster robust standard errors give false positives on cross-level interactions?

One of the most easy traps to fall into in quantitative social science is fooling yourself (and your statistics …

Read On →

Why is this the answer?

In my previous post I showed that the minimum inverse-probability weight in a simple random sample should be equal to …

Read On →

Farage's Brexit Party Announcement and the 2019 Campaign

John Curtice, Stephen Fisher, and Patrick English made headlines recently for claiming that Nigel Farage’s decision to …

Read On →

A simple function for creating dummy variables in R

As someone interested in economic voting, I create a lot of dummy variables. Most often, I have to use self-reported …

Read On →

The minimum post-stratification weight in a simple-random-sample equals the response rate

As my first post on the File Drawer, I wanted to share a survey statistics result that surprised me and that I haven’t …

Read On →