Alex Reinhart

I’m an assistant teaching professor of Statistics & Data Science at Carnegie Mellon University. I work on statistical models to understand and predict where events occur, while also helping lead the Teaching Statistics Group study how students learn statistics and how we can teach better.

You should read my book Statistics Done Wrong, the woefully complete guide to the many ways that statistics is abused in science. Now available in print, published by No Starch Press!

You might be interested in

• My publications and research,
• My presentations,
• My courses taught,
• Or my CV.

You can contact me at areinhar@stat.cmu.edu or find me in my office, Baker 232K.

Notebooks

I keep extensive online notebooks on random topics of interest to me. Some recent updates:

• Privacy and surveillance – April 24, 2023
• Algorithmic fairness – April 16, 2023
• Causality – December 2, 2022
• Regression – September 2, 2022
• Dimensional analysis in statistics – August 4, 2022
• Survey sampling – June 14, 2022
• Algorithmic due process – May 7, 2022
• The many-hands and many-causes problems – January 18, 2022
• Machine learning and law – January 10, 2022
• Statistical misconceptions – August 10, 2021
• …and a bunch more.

Blog

Recent posts from the refsmmat report:

• Powering accelerometers for fun, profit, and field recording – February 12, 2022
  What if you could hear the inside of a watermelon? Building an ICP signal conditioner for accelerometers and measurement microphones.
• Using statistics and cognitive science to understand how students learn statistics – February 6, 2020
  For the past two years, we’ve been conducting research on how students learn statistics. What have we learned so far?
• Books of 2017 – February 2, 2018
  Brief reviews of a selection of books I read in 2017.
• Books of 2016 – February 26, 2017
  Brief reviews of a selection of books I read in 2016.
• R’s Lists and its Detestable Dearth of Data-Structures – September 12, 2016
  A deep dive into R’s inner workings.

All posts · RSS

Other writing

• Huff and puff (preprint PDF here), the story of Darrell Huff’s lesser-known book How to Lie with Smoking Statistics. Accompanying source documents are available.
• Belief-Sustaining Inference, published in the Proceedings of SIGBOVIK 2014. Why data is bad for you, and for statistics.
• Strategic Buzzword Council closed due to budget cuts.
• Gnome Physics – how all of modern physics can be explained through gnomes.
• The Illegal Operation: when nerds write poetry.
• Independence Day 2.0: aliens discover the dark side of humanity.
• Breaking News: Strategic Hyperbole Reserves Depleted.

Old projects

• ScienceForums.net, one of the world’s largest online discussion forums focused on science and science news.
• The Regressomatic, an interactive demonstration of linear regression diagnostics.
• jsphys, a Javascript-powered special relativity simulator for classroom use.
• madman, a Python-based Mad Libs generator that uses Markov chains to generate rather alarming texts. Provide a perfectly sensible text file and madman makes it less sensible.
• seuss, a Python-based poetry generator that uses Markov chains to create humorous and surprising poems.