Ryan Briggs

We have a new version of this paper out. The headline results are the same—political science must filter results heavily for statistical significance—but we've added many extensions and rewritten much of it in response to feedback (thank you!). A quick thread on updates 👇

Ryan Briggs@ryancbriggs.net

I have a new paper. We look at ~all stats articles in political science post-2010 & show that 94% have abstracts that claim to reject a null. Only 2% present only null results. This is hard to explain unless the research process has a filter that only lets rejections through.

Figured out how to make soft serve swirls. This one is ube and black sesame.

22 Sept

Reminder to Toronto people: It is mooncake season. Get your mooncakes.

Ryan Briggs reposted

🇨🇦 CPSA ACSP 🇨🇦

@cpsa-acsp.bsky.social

17 Sept

The Call for Proposals #CPSA_ACSP27 is live. ⏰ Nov 10 2026 Meet the colleagues behind next year's programme — and send them your work! 📍 UBC-Vancouver-CANADA|May 26–28 2027 mycpsa-cpsa-acsp.ca/cfp/cfp_index @traceyraney.bsky.social @svallejovera.bsky.social @ryancbriggs.net @randybesco.bsky.social

14 Sept

All academics fantasize of either opening a bakery or escaping to a remote cabin in the mountains

10 Sept

When I was visiting my mom she gave me some bottles of alcohol that she was never going to drink, and I just opened this one and turns out it’s maple syrup

6 Sept

Nice place to take an APSA panel

4 Sept

This is clever

Institute for Replication@i4replication.bsky.social

🧵 New DP from I4R co-director Lester Lusher, with Sonia Ale, Md Shafiqui Islam, Huda Osman & Jacob Stenstrom (Pitt): "When Samples Shape Significance." What happens when you re-estimate published results using fresh draws of the same data? Some evidence from Google Trends!

docs.google.comAlex Usher email to Brian Lee Crowley(text of an email sent to Brian Lee Crowley of the MacDonald Laurier Institute on September 2, 2026) Dear Brian, I am writing to you and your board with respect to the publication “The equity excelle...

4 Sept

It has been embarrassing reading this and similar critiques, because they accuse the report of not doing things that it manifestly does do (like adjusting for time to accumulate citations). I hope everyone involved is more careful in their regular work. You hurt your causes by flubbing this hard.

Alex Usher@alexusherhesa.bsky.social

Here is my letter to the MacDonald Laurier Institute on the subject of their appalling research. Use as you will. I recommend sending something of your own, though. And if you know anyone associated with MLI - and Board directors, say - send it to them, too.

4 Sept

If you were mad about the original report then you should also be mad about garbage criticism! Everyone saying “this is good” is wrong. It is in fact bad. The more you think the subject matters, the more upset you should be about factually incorrect criticism that makes the critical side look stupid

1 Sept

The haskaps finally came in at my parent’s place! We made jam and haskap sour cocktails and a haskap pie is in the oven.

18 Aug

It's quite neat that the 1954 Salk Polio trial in the US had both a "matched" observational arm and a conventional RCT arm. There is an R package with the relevant data (PolioTrials in HistData) as well as an article about it by Paul Meier (webpages.math.luc.edu/~mgb/courses...).

18 Aug

The main takeaway is that both arms produced the same signed effect, but the RCT estimate for the vaccine's effectiveness was about 10pp higher than the observational one. It makes an excellent intro teaching resource for a course on causal inference.

Ryan Briggs reposted

Michael Clemens

@mclem.org

aiddata.orgBefore reimagining development data, remember what we’ve learnedWhy we’re publishing a free, searchable, and ungated archive featuring a quarter-century of USAID evaluation reports.

Incredible. When E—n M—k & his crew obliterated USAID in a weekend, they deleted a vast trove of publicly-funded knowledge: the Development Experience Clearinghouse. A Canadian high school student (!) happened to have downloaded the entire thing for a project. AidData has now posted it for all.

asteriskmag.comWhy We Like ThingsWhat Silicon Valley gets wrong about taste.

I enjoyed this Asterisk article about taste: asteriskmag.com/issues/15/wh... My favourite little-known example of outstanding taste is cookbook editor Maria Guarnaschelli. Just while working at Norton her authors won 13 James Beard awards & at least 13 IACPs! Here are some of the books she edited:

* All About Roasting (2011) Molly Stevens * Land of Plenty (2003) Fuchsia Dunlop * The All New, All Purpose Joy of Cooking (1997) Irma S. Rombauer, Marion Rombauer Becker, and Ethan Becker * Classic Indian Vegetarian and Grain Cooking (1985) Julie Sahni

* The Bread Bible (2003) Rose Levy Beranbaum * The Pie and Pastry Bible (1998) Rose Levy Beranbaum * My Bread (2009) Jim Lahey * Ratio (2009) Michael Ruhlman

asteriskmag.comWhy We Like ThingsWhat Silicon Valley gets wrong about taste.

* The Zuni Café Cookbook (2002) Judy Rodgers * The Splendid Table (1992) Lynne Rossetto Kasper * Gran Cocina Latina (2012) Maricel E. Presilla * Rick Bayless’s Mexican Kitchen (1996) Rick Bayless * Mexico One Plate at a Time (2000) Rick Bayless * The Food Lab (2015) J. Kenji López-Alt

* The Cake Bible (1988) Rose Levy Beranbaum * Classic Indian Cooking (1980) Julie Sahni * All About Braising (2004) Molly Stevens * Every Grain of Rice (2012) Fuchsia Dunlop * BraveTart (2017) Stella Parks * Liquid Intelligence (2014) Dave Arnold

27 July

I’m late to this but the ninja creami is viral for good reason. It’s wild how good of a product you can get out of it, including vegan ice cream (almond milk + coconut cream).

maps.app.goo.glmaps.app.goo.gl

26 July

The food court in the Mississauga flea market has incredible food. We got Malaysian but just in general it seems very strong. maps.app.goo.gl/KkoR12QxjZsN...

Ryan Briggs reposted

Paul Nightingale

@paulnightingale.bsky.social

11 July

We are seeing similar at RP. It's not just AI. Also a huge increase in Chinese social science. Which is often very good. But AI slop is flooding the system. One author with 30 submissions in a month... now they "could" be super productive.... or it could be AI slop.

Ryan Briggs@ryancbriggs.net

Spoke to an editor of a good social science journal who reported submission volume is now 400% higher than during Covid. He entirely blames AI.

11 July

Spoke to an editor of a good social science journal who reported submission volume is now 400% higher than during Covid. He entirely blames AI.

2 July

I think he deserved more than a dark corner in the Guinness museum

22 June

Homemade ice kacang. I probably made lots of moderate execution errors, but it was good enough that the kids happily crushed it.

20 June

I won’t get into the data or analysis here (yet), but I’m very confident that lower ranked journals do not publish more null results than higher ranked ones. I looked across 11 disciplines and all have this (lack of a) pattern.

daniellakens.blogspot.comEvaluating Dr. Cuddy’s Claim that the Debunking of Power Posing is a MythIn this blog post I will analyse the arguments that Dr. Amy Cuddy provided in a blog post “The "Power Posing Was Debunked" Myth: What the Re...

10 June

This was a fun paper. We analyze results from a global survey of development studies (DS) profs and find that the income group of their country doesn't explain much about their views on DS. Instead, their discipline is huge.

Development Studies Association @devcomms.bsky.social

In a new @eadi.bsky.social reflection paper, @ryancbriggs.net & @andypsumner.bsky.social, @kings-sga.bsky.social explore what drives deep disagreements in development studies—and point to the “disciplinary baggage” of PhD training. Read more: buff.ly/ASH1bL9

Ryan Briggs reposted

Daniel Lakens

@lakens.bsky.social

25 May

New blog post: Evaluating Dr. Cuddy’s Claim that the Debunking of Power Posing is a Myth. daniellakens.blogspot.com/2026/05/eval... On an AI generated description of a non-existent study, incorrectly citing findings from studies, and the importance of scientific criticism.

25 May

You’ve got to be clafoutis maxxing

23 May

@prachisrivas.bsky.social you would love the number of people wearing Habs jerseys at this Toronto Blue Jays game

In the Fall I'll be teaching a new MA-level methods course entitled "Applied Statistical Evaluation of Development Projects". It will be 12 weeks, in R, and aimed around RCT evaluations. This is a draft outline. What am I missing? What seems redundant?

It has to accommodate people that are new to R and lack much background in statistics, while also absorbing people who are much more technical (this is hard). The goal is to make students into very adept consumers of RCT evidence more than evidence producers.

I should probably add attrition and handling it to week 8? Maybe week 9 is too thin? I dunno. this is all just what I jotted down in a quick brainstorm

forum.effectivealtruism.orgGlobal Health Charity Founders on GiveWell, Evidence Action, and M+E — EA ForumThe following is a lightly edited and anonymized transcript of a discussion among charity founders, researchers, and funders in Ambitious Impact’s Sl…

An interesting EA forum post that is an anonymized slack chat about GiveWell and EA-style CE analysis: "across EA cost-effectiveness analyses, much more typically goes toward estimating the effect than estimating the costs." forum.effectivealtruism.org/posts/ru3wyS...

21 May

I'll be teaching a grad methods class focused on experiments in the fall for a development studies dept. I'm looking for published development RCTs. Ideally they would have: 1. replication packages in R, or 2. fairly simple code I could rewrite in R. Anyone have recommendations?

Ryan Briggs reposted

Lauren Gilbert

@lgilbert.co

indevelopmentmag.comJobs – In Developmentevidence and argument for a developing world

I'm hiring for In Development's first not-me hire. Help shape the future of development journalism; edit pieces that do serious cost-effectiveness analysis AND have jokes. Up to $90k USD, fully remote worldwide. indevelopmentmag.com/jobs/

Ryan Briggs reposted

Ethan Mollick

@emollick.bsky.social

21 May

We can estimate the resource cost of solving the Erdos problem. The calculations below seem reasonable, so using the best public estimates we have, it took 0.6–6.3 kWh of electricity & 3–31L of water Equivalent to < 3 almonds worth of water & the electricity equivalent of 2-20 miles of EV driving

Ryan Briggs@ryancbriggs.net

Any idea how many cheeseburgers worth of water it took to solve this?

openai.comAn OpenAI model has disproved a central conjecture in discrete geometryAn OpenAI model solved the 80-year-old unit distance problem, disproving a major conjecture in discrete geometry and marking a milestone in AI-driven mathematics.

“Stochastic parrots”

Timothy Gowers@wtgowers.bsky.social

OpenAI's claim that this is a central conjecture in discrete geometry is not an exaggeration. This will I think be looked back on as the first time that AI solved a major mathematics problem (defined as a problem that all experts in some subfield had thought about). openai.com/index/model-...

Ryan Briggs reposted

Ethan Mollick

@emollick.bsky.social

June 2024: The latest general-purpose LLMs could not count the r's in strawberry. July 2025: The latest general-purpose LLMs get gold in the International Math Olympiad. May 2026: The latest general-purpose LLM solve an 80 year old problem, one of the "best-known questions in combinatorial geometry"

Ryan Briggs reposted

Eris ⌬

@eriskii.net

OpenAI has used a "general purpose reasoning model" to disprove that the square grid type construction is the best solution to the planar unit distance problem. I am sure the "AI is completely useless" crowd will now change their ways, right?

Ryan Briggs reposted

It's Jamie

@sjamieit.bsky.social

The City of Ottawa posted the raw speed data from before and after Doug Ford banned speed cameras and the effect is so obvious it's not even necessary put a line marking when that happened. This chart shows the percentage of drivers going 15km/h or more over the posted limit.

Ryan Briggs reposted

Julia M. Rohrer

@dingdingpeng.the100.ci

Jeremy right now

Jeremy Labrecque@jeremylabrecque.bsky.social

A slide from this talk I'm giving on Monday. The conclusion will be a set of questions that one must be able to answer to ask associational research questions. The ball is in your court association people!

Ryan Briggs@ryancbriggs.net

5. Finally, we discuss an editorial rule of "selection on precision," where one selects results on the width of confidence intervals rather than p < 0.05. It has fairly good properties for the normal linear model, but outside of that it introduces bias in coefficients (like SOS).

This is the new paper: osf.io/preprints/so... Thank you to everyone who gave us feedback!

Ryan Briggs@ryancbriggs.net

3. People that did experiments also told us that they thought that null results were often indicative of a failure to manipulate. We worked out a little model with Bayesian updating to show that this point is mostly misguided. If you want evidence of dosage, you really should collect it directly.

4. Everyone told us that the vibe of the last article was that "all nulls should be published." We now have a section describing the kind of null results that should not be published. However, many things that make a null result unpublishable should also make a sig result unpublishable.

Ryan Briggs@ryancbriggs.net

1. We have a better framework for calculating benchmark abstract shares under no selection on significance. We show how our results vary across wide parameter sweeps. I like the transparency of this new approach. All reasonable combinations of parameters lead to large degrees of SOS.