Skip to primary content

Statistical Modeling, Causal Inference, and Social Science

Main menu

  • Home
  • Authors
  • Blogs We Read
  • Books
  • Sponsors

Post navigation

Be careful what you control for . . . you just might get it! Luck or knowledge?

Blog on applied probability modeling

Posted on August 26, 2011 9:11 PM by Andrew

Joseph Wilson points me to this blog on applied probability modeling. He sent me the link a couple months ago. If he’s still adding new entries, then his blog is probably already longer-lasting than most!

This entry was posted in Miscellaneous Statistics by Andrew. Bookmark the permalink.
  • Art
  • Artificial Intelligence
  • Bayesian Statistics
  • Causal Inference
  • Decision Analysis
  • Economics
  • Jobs
  • Literature
  • Miscellaneous Science
  • Miscellaneous Statistics
  • Multilevel Modeling
  • Obituaries
  • Papers
  • Political Science
  • Public Health
  • Sociology
  • Sports
  • Stan
  • Statistical Computing
  • Statistical Graphics
  • Teaching
  • Zombies
  1. Erling on Blokjutsu: a macOS image generator/editorNovember 18, 2026 3:47 AM

    Love the occasional RPG content! Only blog that serves two separate interests of mine (though only occasionally). If you end…

  2. David, a Bostonian in Tokyo on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 18, 2026 3:11 AM

    I wrote a one-page (1000-word) newsletter to go with my new year's cards I send out every year to a…

  3. Jessica Hullman on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 10:43 PM

    Vipul - I'm not sure I understand what you mean by bring the two factors together. By priors I meant…

  4. Joshua on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 10:38 PM

    Vipul - That's an interesting comment I think of writing as thinking in much the same way that discussing or…

  5. Jessica Hullman on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 10:03 PM

    It's the same thing. As the post says, Pangram considers a false positive any segment being classified as AI.

  6. Michael Murphy on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 6:07 PM

    Andrew wrote: "Do you have evidence for your claim that the accusations are in most cases obviously false? This doesn’t…

  7. Dale Lehman on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 5:40 PM

    The post states "The result is a dent in our collective understanding of basic binary classification." But, as you say,…

  8. Andrew on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 5:39 PM

    Michael: 1. Do you have evidence for your claim that the accusations are in most cases obviously false? This doesn't…

  9. Michael Murphy on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 5:23 PM

    So this begs the question of why so many of these so-called scandals involve black writers? Again, the accusations are…

  10. Michael Murphy on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 5:20 PM

    Because they are defamatory and the careers of people who don't get much anyway are being damaged. Again, any writer…

  11. Vipul Vivek on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 5:15 PM

    On the broader question discussed in the last section: most of the social media discussion about „writing is thinking“ is…

  12. Michael Murphy on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 5:13 PM

    Pardon me. This is correct. General point holds.

  13. Dale Lehman on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 5:04 PM

    sorry, I wasn't thinking.

  14. Andrew on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 5:03 PM

    Michael: Just speaking in general terms, I'm not happy with any sentence that begins, "These accusations are defamatory and frankly…

  15. Alex Reinhart on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 4:59 PM

    Your characterization doesn't seem to be accurate: the Caine Prize website says "The writer’s process started in April 2025 from…

  16. Jörg Wittkewitz on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 4:42 PM

    Sure thing, the I in LLMs called AI came from human labelers and RLHF work. Without these guys today no…

  17. Michael Murphy on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 4:39 PM

    So there are some basic facts at issue here. The most important one is that most of the accusations levelled…

  18. Andrew on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 4:33 PM

    Not anthropomorphic. Parrotomorphic.

  19. Ewout ter Haar on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 4:18 PM

    The argument was too anthropocentric? As in assuming that man is the measure of all things? I think I'd agree…

  20. Andrew on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 4:15 PM

    Jorg, I think that when people speak of LLMs they're not just talking about the string prediction but also about…

  21. Joshua on 2026 election conspiracies haven’t breached the Vermeule threshold.November 17, 2026 4:05 PM

    J. J. - Yes. John's comment makes perfect sense. I was being sarcastic. Look at the link and the associated…

  22. Jukka on Algorithmic deepfake scienceNovember 17, 2026 4:03 PM

    "Such papers could include plausible computer-generated images and data, along with chatbot-generated text describing nonexistent experiments." Why "could"? It is…

  23. addi on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 3:41 PM

    i think the entire: lets test on pre LLM text thing is unserious. do we know what pangram was trained…

  24. Jörg Wittkewitz on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 3:25 PM

    LLMs are nothing but string predictors. So all ai detection is just measuring surprise in terms of a next word…

  25. Bob Carpenter on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 3:01 PM

    In order to talk about false-positive and false-negative rates, you need to actually set up a classifier. So even if…

  26. Bob Carpenter on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 2:54 PM

    I would've said her stochastic parrots argument suffered from being too anthropomorphic rather than being woke.

  27. Joshua on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 2:25 PM

    Yes, sorry. I intended to reference that in my comment, but forgot.

  28. Joshua on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 2:23 PM

    From Kriss: "attractor in the fabric of the discourse" and "discursive rearmament." Hmmm. It would be interesting to see what…

  29. Jessica Hullman on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 2:21 PM

    Yes, like I mention in the post, there was an early study by James Zou and others that claimed bias…

  30. Joshua on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 2:11 PM

    As it happens, I recently submitted a question into Claude for names of programs to detect AI text, and the…

  31. Anonymous on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 2:10 PM

    Josh: That's interesting. I checked the comments on the post I linked, and indeed, several recent comments make the same…

  32. J. J. Ramsey on 2026 election conspiracies haven’t breached the Vermeule threshold.November 17, 2026 2:09 PM

    I'm not sure what your objection to John's comment was. My takeaway was that Texas' blue shift wasn't due to…

  33. Colin on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 2:01 PM

    People have investigated this, using data from freshly-scanned books that were unavailable to use as training: https://veryfineprint.substack.com/p/scanning-for-pangram-errors Furthermore, others have…

  34. Seth Finkelstein on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 1:38 PM

    First, "racist" is a really the wrong word. The correct term is "colonialist" :-), though that's out of favor now…

  35. Josh on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 1:32 PM

    To anonymous who wrote "Also, I found a false positive." The link you posted is to an article published in…

  36. Kevin on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 1:27 PM

    This is frequently brought up and just misunderstands how these models work. You do not train a classifier like this…

  37. Anonymous on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 1:18 PM

    Corrected: "I bet she is still wrong on her broader points, but..." Also, I found a false positive: Pangram says…

  38. gec on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 12:49 PM

    I appreciate this post because, as someone with a strong prior skepticism regarding anything that claims to "detect AI", you…

  39. Anonymous on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 12:27 PM

    What do you think about Bender’s criticism on out-of-domain generalization claims? FWIW, I’ve seen the audit results being used to…

  40. Anonymous on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 12:18 PM

    Dale, she says the case studies talk only about false negatives and false positives separately (as if they have no…

  41. Bob Carpenter on Postdoc opportunity! Here with me at Columbia! Cutting-edge Bayesian analysis and causal inference: Statistical Tools for Generalizing in Education Research.November 17, 2026 12:00 PM

    Starting with a plan is almost always a plus, but working with Andrew is always much more open ended, free…

  42. Jessica Hullman on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 11:59 AM

    My name's not Jennifer!

  43. Tim F on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 11:54 AM

    Something I dislike about the Pangram discourse is that people are reporting low false positive rates on pre-2022 text as…

  44. Will on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 11:49 AM

    "Resisting AI as a solution to anything. Finally, I think a lot of people resent the imposition of AI in…

  45. Dale Lehman on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 11:49 AM

    I guess I don't understand your (and Jennifer's) complaint about Bender. I didn't think what she wrote was wrong. You…

  46. Dale Lehman on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 11:24 AM

    I know nothing about Pangram but your post makes me curious about the binary classification aspect. The classification depends on…

  47. Anonymous on The 2026 Pangram social justice misinformation wars: Collective understanding of binary classification takes a hitNovember 17, 2026 10:59 AM

    Bad take from Bender. That’s rare, she’s usually pretty good. Wokeness strikes again…

  48. Bob Carpenter on Blokjutsu: a macOS image generator/editorNovember 17, 2026 10:53 AM

    I think the statistics posts are the most fun. So only very rarely. I've been meaning to do one on…

  49. AAAnonymous on A plan to replicate the Loftus and Palmer (1974) experiment on false memoriesNovember 17, 2026 9:54 AM

    Things have changed indeed....there is also mention of the OSF in the proposal but they have since stopped with a…

  50. Andrew on A plan to replicate the Loftus and Palmer (1974) experiment on false memoriesNovember 17, 2026 9:39 AM

    Things have changed . . . Scientific American is no longer a high-profile outlet!

Proudly powered by WordPress