Herding, polls of polls and Bayes’ theorem
You have probably heard of that outlier Iowa poll that gave a surprisingly large advantage to Harris in the 2024 election. Related to this is that is what forecasters call…
When do simple methods outperform regression?
This is a nerdy post of little interest to the layman. There's this old paper that sometimes comes up: Dana, J., & Dawes, R. M. (2004). The superiority of simple…
So you want a shorter test
There are 1000s of tests, scales, and assessments that have been made in the social sciences and in life in general. Sometimes, these are far too long. Take, for instance,…
The OECD fallacy
I tweeted this funny (?) rebuke at Richard Hanania earlier: However, Twitter demanded that I explain my words: Alas, the fact is that: But to be more serious. The fact…
Using the Binomial Effect Size Display (BESD) to understand correlations
When I talk science to non-scientists, I often get questions about how to interpret sizes of correlations or other metrics. While one can give rough guidelines invariant of the topic…
Measurement error and group differences explained
I have often talked about how measurement error impacts correlations between two variables. Here we are talking about classical measurement error, which is theoretically understood as adding a normally distributed…
Convenience samples are fine because interactions are mostly not real
Attacking a study for using a non-representative, typically online convenience sample is common practice. However, for most studies, this is wrong-headed and here I want to argue why that is…
Brief introduction to test bias in mental testing
Although Meng Hu has recently provided a very thorough review of test bias methods, I thought it would be helpful to have a more visual, intuitive approach for those not…
Interactive statistics: breeding’s equation AKA regression towards the mean
One of the things I did to teach myself R programming was to create a bunch of interactive statistics tools written using the awesome R Shiny package. These are still…
Economics research is mostly untrustworthy because economists cheat so much with the statistics
I recently posted a book review of Garett Jones' 10% less Democracy. My main beef with the book is that it relies on lots of studies with dubious p values,…