vincentab.bsky.social
Prof. Most tweets about R. “Polisci, it’s all about what’s going on.”
http://arelbundock.com
581 posts
5,672 followers
1,333 following
Getting Started
Active Commenter
comment in response to
post
lol. 🤞
comment in response to
post
That's far too sweeping a critique. There are many good use cases for surveys even in low information environments.
comment in response to
post
I like that one. It's one of the clearer treatment of the relationship between collapsibility and generalizability/transportability.
comment in response to
post
Oh cool. Still, not sure I should impose on a stranger to explain their work to a neophyte like me. 🤷
comment in response to
post
Right. But in this Popperian account, severity is not affected by p-hacking, because the friendly-hostile community of critical scientists can always determine the "sincerity" of a test. If that's the world where pre-reg doesn't matter, I don't feel super comforted.
comment in response to
post
Not sure I get it. Is the key argument that pre-reg doesn't matter (a) in the Popperian view if science is self-correcting via a community of friendly-adversarial scientists, and (b) in the Mayo view when people don't follow their pre-reg?
comment in response to
post
lol yeah, I had forgotten about that! It was my "learn python" project in grad school. Feels like forever ago
comment in response to
post
And a real downer for new (better software). Status quo bias is going to be massive.
comment in response to
post
Félicitations!!
comment in response to
post
Yeah, this is a very cool graph.
comment in response to
post
Yep. I need to release `tinytable` before the next version of `modelsummary`, but you can install both from Github now. Sorry for the issue. Please feel free to open a bug report on Github next time you see something weird like that.
comment in response to
post
No no. The nudge was very useful. I completely forgot about this when I updated the site.
comment in response to
post
Sorry about that. Link is back up here: arelbundock.com/posts/dt_tb_...
comment in response to
post
Wow that Danish looks craaazy. (And I'm happy you got to see that show! Must have been cool.)
comment in response to
post
it's a good one
comment in response to
post
I like this frame a lot! And I'll add another incentive category (what I personally care most about): lowering the cost of good science.
comment in response to
post
Right. I'm sure this happens. In such cases, I feel one benefit of prereg is that it turns a routine behavior into something that feels obviously unethical. And I think many people will feel bad / refrain. But maybe I'm too optimistic. Again, I'm super interested to see more systematic evidence.
comment in response to
post
Interesting case, and I appreciate your work in that space! But I wonder if the prior you are talking about is one about the bias of norm entrepreneurs, or about the effect size of a given intervention.
comment in response to
post
I'd love to see more systematic evidence! FWIW, I've been in the room many times when audience members recommended some weird fishing/phacking trick, and the author said "I wish I could but we didn't preregister that." That's obv weak, unsystematic evidence, so the prior is still doing work for me
comment in response to
post
Tried to guess what this was about by looking at your Google Scholar, and learned you wrote that Shapley paper from a while ago. That was fun stuff. I remember forwarding it to a bunch of people.
comment in response to
post
I think there's a good chance it will work out of the box, but please open an issue on Github if it doesn't.
comment in response to
post
Glad to hear this might be useful! (Robin Denz deserves ~all the credit for this awesome vignette.)
comment in response to
post
That *is* a cute typo. Thanks for taking a look and reporting!
comment in response to
post
Unfortunately, coxme is not supported because of a limitation in that package's predict() method. See here for discussion: github.com/vincentarelb...
comment in response to
post
🤫. (But Robin did essentially all the work on the vignette. That was amazing.)
comment in response to
post
yeah, that looks amazing
comment in response to
post
This was super fun and informative. Thanks for writing and posting!
comment in response to
post
Love this!!
comment in response to
post
😰
comment in response to
post
Excited to listen to this. Looks like you've got the full Twitch Stream setup!
comment in response to
post
Some people think it's fine to truncate symmetric delta method CIs, but that's obv a hack. Using inferences() to bootstrap or simulate CIs is an alternative.
comment in response to
post
This can happen sometimes because 95% CIs are built symmetrically: [B-1.96*SE, B+1.96*SE]. In predictions() we use backtransformation to avoid this issue, but that strategy is not available/straightforward everywhere.
comment in response to
post
Thanks! I *really* appreciate bug reports. Github is usually better because I sometimes miss/forget social media.
I agree that 1 is a better default w/ `hypothesis=ratio ~.`
The default is changed in version 0.25.1.7 on Github. The new version will be published to CRAN in the next few weeks.
comment in response to
post
Are you talking about an "agenda" or a "platform"? Reforming the Senate might be good policy but not a debate/vote winner.
comment in response to
post
I watched The Rock (again) yesterday. (sorry had to share)
comment in response to
post
I mis-wrote above. The distinction in this paper is different: "Treatment effect heterogeneity" vs. "Causal moderation." We might be able to estimate that treatment X has stronger effects on people with kids (former) even if we don't know that the effect is strong *because* they have kids (latter).
comment in response to
post
How does this language map on to the "Treatment Effect Modification vs. Causal moderation" in Section 2.2 academic.oup.com/jrsssa/artic...
comment in response to
post
As Konstantin notes, a good first step would simply be for code to be available to reviewers by default on submission. I would certainly look at it if provided. This should be ~0 additional burden on authors: if the code is not clean enough to be shown, the paper is not ready for submission anyway.
comment in response to
post
Like: implement new feature X by adding an argument to a function, which requires modifying 5 other functions in a codebase. I don't care about language or toolkit. Just general workflow.
comment in response to
post
I've had great success with discrete tasks like "write a SQL query to do X" but have mostly failed with things involving several functions. I suspect that my problem is a high-level "approach" issue, so I'd like to observe someone else's successful workflow to figure out what I do differently/worse
comment in response to
post
Same for me so far. I suspect that part of it is I'm not yet good enough at thinking of small discrete chunks.
comment in response to
post
Let's say "existing R package"-large.