10th anniversary of “Statistical Modeling, Causal Inference, and Social Science”

Richard Morey pointed out the other day that this blog is 10 years old!

During this time, we’ve had 5688 posts, 48799 comments, and who knows how many readers.

On this tenth anniversary, I’d like to thank my collaborators on all the work I’ve blogged, my co-bloggers (“This post is by Phil”), our commenters, Alex Tabarrok for linking to us way back when, and also the many many people who’ve pointed us to interesting research, interesting graphs, bad research, bad graphs, and links to the latest stylings of David Brooks and Satoshi Kanazawa.

It’s been fun, and I think this blog has been (and I hope will remain) an excellent communication channel on all sorts of topics, statistical and otherwise. Through the blog I’ve met friends, colleagues, and collaborators—including some such as Basbøll and Palko whom I’ve still not yet met!—; I’ve been motivated to think hard about ideas that I otherwise would’ve encountered; and I’m pretty sure I’ve motivated many people to examine ideas that they otherwise would not have thought seriously about.

The blog has been enlivened with a large and continuing cast of characters, including lots of “bad guys” such as . . . well, no need to list these people here. It’s enough to say they’ve provided us with plenty of entertainment and food for thought.

We’ve had some epic comment threads and enough repeating topics that we had to introduce the Zombies category. We’ve had comments or reactions from culture heroes including Gerd Gigerenzer, Judea Pearl, Helen DeWitt, and maybe even Scott Adams (but we can’t be sure about that last one). We’ve had fruitful exchanges with other researchers such as Christian Robert, Deborah Mayo, and Dan Kahan who have blogs of their own, and, several years back, we launched the internet career of the late Seth Roberts.

Here are the titles of the first five posts from our blog (in order):

A weblog for research in statistical modeling and applications, especially in social sciences

The Electoral College favors voters in small states

Why it’s rational to vote

Bayes and Popper

Overrepresentation of small states/provinces, and the USA Today effect

As you can see, some of our recurrent themes showed up early on.

Here are the next five:

Sensitivity Analysis of Joanna Shepherd’s DP paper

Unequal representation: comments from David Samuels

Problems with Heterogeneous Choice Models

Morris Fiorina on C-SPAN

A fun demo for statistics class

And the ten after that:

Red State/Blue State Paradox

Statistical issues in modeling social space

2 Stage Least Squares Regression for Death Penalty Analysis

Partial pooling of interactions

Bayesian Methods for Variable Selection

Reference for variable selection

The blessing of dimensionality

Why poll numbers keep hopping around by Philip Meyer

Matching, regression, interactions, and robustness

Homer Simpson and mixture models

(Not all these posts are by me.)

See you again in 2024!

13 thoughts on “10th anniversary of “Statistical Modeling, Causal Inference, and Social Science”

  1. Andrew: Let me be the first to officially congratulate you (in this forum) for a decade of blogging! You’re the leader in stats blogs hands down, and one of my main guides for how to do it–even though I could never keep up with you in my slow, ponderous, philosopher’s blog of 3 years. Thank you for mentioning me, even though I’m sure I’ve sometimes had to take the role of one of the “bad gals”.

    • Thanks, Mayo.

      The blog world has changed in the past decade—in the early 2000s, it seemed that there were a lot more back-and-forth between blogs.

      Not so much now. I wonder where all the bloggers all went. I guess some of them moved up in the world and are now professional journalists, while others got what they wanted out of blogging and retired, while still others had viewed blogging as a method of professional advancement and stopped doing it after awhile.

      Now there’s a lot of tweeting but, grumpy old man that I am, I don’t think it’s the same thing. A tweet is a reaction, a link, or a quip, but I don’t think it takes the place of the sort of open-ended exploration you can get in a blog. I’m just sad that, outside the field of economics, there’s so little in the way of conversations between blogs.

      I also wonder what I would’ve been doing with my time had blogging never existed. Lots of emailing and posting to listservs, I guess. And maybe I would’ve written a few more books by now.

      • Andrew: At some point I’d like to go back to interspersing between the blog and an on-like publication as with the RMM volume, and “U-Phils”, but not til I finish my book. It is time-consuming, even as I’ve pulled back my posting,and the novelty can wear off, I guess. One of the justifications for me was posting pieces of the book, because it forced me to finish sections, and gave me quick feedback. But now I’m keeping most of it under wraps, but not all. When it’s done, I’ll likely focus on running workshops and conferences and the blog can connect with that.

  2. Congratulations. This blog is –in econspeak– a public good that we can all enjoy thanks to Andrew’s generosity. I have been doing more multi-level modeling in the last few years, much of it Bayesian. And visiting this blog has now become a daily ceremony (together with marginalrevolution). I particular enjoy the posts on “bulshit-detection”, i.e. when Andrew points out sloppy methods (sloppy thinking!) in published papers and popular press articles. I particular favorite of mine is when he “doesn’t understand” structural equation modeling and thing like that. I smile and discreetly fist-pump in front of my laptop.

  3. Congratulations! Your steady stream of interesting and informative posts is remarkable. I’m also very fond of the comments here (unlike nearly every other place in the internet), which I think you nicely set the tone for. I look forward to the next ten years — thanks!

  4. Andrew:

    I also wonder what I would’ve been doing with my time had blogging never existed. Lots of emailing and posting to listservs, I guess.

    Well, that would have been a shame. Blogging is a great use of your time:

    blogging is a communication pattern that optimizes for the amount of awareness and influence that each keystroke can possibly yield. source

    So next time any reader wants to write an email about some stats problem they should do this little calculation.

    • Fernando:

      I clicked on your links.

      I agree with the first one; indeed, I did have repetitive stress injury several years ago myself and, as I’ve written elsewhere, one of my reasons for blogging is to reach more people than would be reached by email.

      The second link seems just wrong to me, it seems to demonstrate the sort of deterministic reasoning and mindless attention-seeking that I hate. Although maybe I’m missing something, perhaps it’s a parody site?

      • Ha. Yes, maybe they should provide a probability interval of the expected number of keystrokes left. But mostly I think it is just a joke, a quick way to make a point.

  5. Congrats and thanks.

    Your blog is a great way for me to keep thinking about applying statistics from many angles and different view points.

    Making a comment requires a level of thought and commitment and even given the possibility that someone might challenge it, its good practice.

    Looking forward to the next 10 years!

  6. Pingback: Friday links: new p-hacking data, grant lotteries, things ecologists will never experience, and more | Dynamic Ecology

Leave a Reply to Hernan Cancel reply

Your email address will not be published. Required fields are marked *