Felix Singleton Thorn

If any data / science / psychology / metascience / etc folks want an invite to the other non-Zuckerberg microblogging site, DM me, I have a few. It’s … mildly better (and at least its CEO doesn’t actively do the opposite of what any competent comms advisor would suggest)

@NickEvershed If you have a spare invite code I would love to have one. (Also love your work).

@Jake_Elder52 I think running lm on an ordered factor is equivalent to adding a n-th degree polynomial, with n = the number of factor levels (by default)? So the results are not going to be the same, the models are totally different. If you expect a linear relationship just model as numeric?

I guess it's close enough that perhaps it doesn't make much of a difference Was just kind curious if this would ever make a difference in treating ordered factor vs. numeric as IV

Someone has a ordered factor with levels [0,20,40,60,80,100]. I recommend they model the IV as an ordered factor with a linear contrast. They ask why they cannot treat as numeric What might differences in est be for an IV treated as ordered factor w/ linear contrast vs. numeric?

@NickEvershed I can’t believe the degree to which the ABC still relies on the worst kind of clickbait headlines. My favourite example (thankfully about a less serious topic) is practically a parody of itself:

@rogierK When errors were found, they were only updated in the group rows, but the p2…p5 data there was inconsistently filled in, so when data was missing you had to fill it in from the individual pt rows. The stuff of nightmares. 3/3

@rogierK If a group had more than 5 pts, they would have multiple group rows. The only way of telling which were “group” rows was if they had pt2…pt5 data, so 6 person groups were… difficult. Worse, this has been done manually, and was internally inconsistent including the pt IDs 2/3

@annemscheel @UniUtrecht Congratulations! They're beyond lucky to have you!

Some news: Just over a month from now I'll be starting as an assistant professor (tenured) in Methodology & Statistics at @UniUtrecht! Hard to contain my excitement about landing at such an amazing place. Looking forward to stealing all your stats teaching materials on here!

Statisticians finally did it! They solved sample size determination!

I'm Hiring! I'm looking to bring in a new data analyst type into a consumer data officer role in my team at @neaminational. Details here and in the thread below, please share with anyone you think might be interested. apply.workable.com/neami-national…

(cont.) RRs came out as-if-not-more-highly rated on all outcomes. Caveats come from the fact that this is a naturalistic study comparing a smallish sample of very early RRs against a matched sample, so this is early indicative evidence and is far from the last word on the topic!

Our paper comparing Registered Reports vs. traditionally published articles is finally out! The tldr is after getting peer reviews of RRs and a matched sample of traditionally published articles (1/2)

Academic Twitter achievement unlocked, finally got a “Want to read the article before retweeting” on a paper I wrote

I just pushed a new version of the metafor package (version 3.0-0) to CRAN. This has been long overdue. This version includes a lot of updates that have accumulated in the development version of the package over the past 14-15 months. I will mention some highlights.

Hi folks, if you feel like critiquing the soundness of papers and predicting whether they will replicate I have the activity for you! Come to a @replicats bushel virtual workshop and assess some papers for us in exchange for $200USD grants!

I am hiring a 5 (3+2) year postdoc on data management and analysis of educational dataset. One critical project is to create a database of publicly available education data but you can pursue your own project too. Come to beautiful Tübingen!! tinyurl.com/58cctbrf

Seems lҽgiᴛ

The curl feature window is now... open. curl.se/mail/lib-2021-…

@drjbeaudry As sad as I am that we probably won’t run into each other as often in Melbourne, I am so thrilled to hear you’ve got such a great position! So many congratulations!