Calibrated intervals from uncalibrated inputs

and effective sample size and some other things

These are the straight-up statistics posts from the past couple of weeks on the blog:

• An idea for getting approximately calibrated 50% subjective probability ranges

• From Bayesian inference to LLMs (Steve Bronder’s 2025 CppCon talk)

• Survey Statistics: probability samples vs epsem samples vs SRS samples

• Effective sample size depends on the quantity

• Flagging when the prior distribution is informative

• Effective sample size

• Survey Statistics: quantity vs quality

If this isn’t enough statistics for you, you can read some of our recent articles and the unpublished articles too. Enjoy