iiinova — where your digital self keeps evolving

Anne Scheel @annemscheel.bsky.social · 23/07/2026

3) Hard publication targets make RRs unpopular if the target can’t be met with RRs alone. If it can, it makes RRs more attractive than at baseline. 4) Competition is the biggest RR-killer in our model. Intense competition makes them widely unpopular & completely unviable in fields with low pace. 11/

110 1

Anne Scheel @annemscheel.bsky.social · 23/07/2026

2) Being able to run many studies before being evaluated (e.g., applying for a job) cancels out most other effects. Researchers with 'slow pace', e.g. in fields with slow or costly data collection, are most affected by increasing/decreasing career returns, publication targets, and competition. 10/

110 1

Anne Scheel @annemscheel.bsky.social · 13/07/2026

Looks like a cool analysis, but note that it excludes preprints that were never published. So this last sentence in the abstract amounts to ‘conditioning on preprints accepted by the scientific community, preprints are good enough to be accepted by the scientific community’.

38913 89

Anne Scheel @annemscheel.bsky.social · 25/05/2026

New estimate for residual bias after study registration just dropped

0122 12

Anne Scheel @annemscheel.bsky.social · 19/05/2026

Doesn’t sound like 300 scans would be too difficult to achieve for someone who’s given *checks notes* >1,3K talks

4200 20

Anne Scheel @annemscheel.bsky.social · 13/05/2026

theories aim to explain and which are thus also crucial for theory building. Here’s a figure from a preprint that I still need to revise that tries to tease these things apart.

140 4

Anne Scheel @annemscheel.bsky.social · 19/03/2026

We used one reaction-time example for illustrative purposes in this paper: academic.oup.com/psychsocgero... It’s very simple but I still quite like it. In general, I think going to applied situations can help a lot with determining SESOIs.

100 1

Anne Scheel @annemscheel.bsky.social · 01/03/2026

Seeing that comments on this are getting a little more political, I want to highlight that my gripe is very much with *the headline* specifically. Neither Alan Milburn nor the author of the article seems to be so seriously confused about the situation. Link to full article: archive.is/pA2lv

120 2

Anne Scheel @annemscheel.bsky.social · 27/02/2026

New newspaper headline for your Intro to Causal Inference lecture just dropped

416633 166

Anne Scheel @annemscheel.bsky.social · 11/02/2026

The comparison is literally the first sentence of your Results section.

110 1

Anne Scheel @annemscheel.bsky.social · 06/02/2026

Thanks, but this strikes me as a very odd and unhelpful definition of “garbage”. If anything, I think one of the things that have hampered progress in psychology is an obsession with “theoretical progress” before establishing any robust phenomena that make mature theories possible in the first place

130 3

Anne Scheel @annemscheel.bsky.social · 24/10/2025

Looks like Study 1E is on executives, N = 44, all others on regular Prolific participants or students. IMO quite misleading to sell this as N=6.4K evidence on “decision makers”. Study 1E asks executives if they’d spend an unspecified (!) amount on sponsored ads on Yelp, where the prior on a causal >

110 1

Anne Scheel @annemscheel.bsky.social · 22/10/2025

It also reminded me of the “stats maven” concept coined by @sanjaysrivastava.com some very long time ago, I believe here? thehardestscience.com/2014/12/04/s...

192 9

Anne Scheel @annemscheel.bsky.social · 22/10/2025

And for my long-held theory, though I admit that this perhaps sounds more cynical than it should (i.e., it happens for benign reasons too and reviewer intuitions can be good/correct even when they’re hard to articulate)

110 1

Anne Scheel @annemscheel.bsky.social · 16/08/2025

“Make it rewarding”: Is there a repository out there that gives you easy access to download and traffic figures? Seems like a no-brainer for encouraging data & code sharing (and high-quality documentation!), yet the OSF’s options for this seem quite limited unfortunately

2150 15

Anne Scheel @annemscheel.bsky.social · 17/06/2025

Worth reading in full. Many important implications here: * Wiley can only publish preprinted papers if final version differs substantially from preprint * After peer review, preprints are no longer preprints (careful with resubmissions!) * Wiley will change journal policies without consulting EiC

44524 45

Anne Scheel @annemscheel.bsky.social · 06/05/2025

To the extent that ChatGPT lets you do stuff faster and turn off your laptop earlier, it looks like it might as well help _save_ energy…

1130 13

Anne Scheel @annemscheel.bsky.social · 30/04/2025

Ah yes, why would open-science ppl be involved with a new journal that’s trying to turn the tide and set new standards re transparency and bias protection in its field, with an EiC (@dangerwhale.bsky.social) who’s been lobbying for RRs and protocol publication for years? It must be the industry $$$

230 3

Anne Scheel @annemscheel.bsky.social · 05/02/2025

The first Swiss edition of Perspectives on Scientific Error kicking off very Swissly indeed #PSE7

0435 43

Anne Scheel @annemscheel.bsky.social · 11/01/2025

I recently learned that Danny Kahneman revolutionised the hiring process of the Israelian army with an improvised personality questionnaire when he was 20 #justKahnemanthings www.moneycontrol.com/news/busines...

090 9

Anne Scheel @annemscheel.bsky.social · 30/11/2024

Cumulative science™️ Paper is here: www.bmj.com/content/339/...

051 5

Anne Scheel @annemscheel.bsky.social · 30/11/2024

Greenberg (2009) is one of my favourite papers, showing “conversion of hypothesis to fact through citation alone” and jarring citation bias in grant proposals. Just struck me that it’s about a literature on beta amyloid—not in Alzheimer’s, but maybe still affected by the awfulness of that research?

44717 47

Anne Scheel @annemscheel.bsky.social · 09/11/2024

Turns out that this conflict/confusion goes back at least as far as Humboldt, who insisted that both the factual and the imagination are crucial for creating scientific knowledge, while others struggled with feeling their imagination threatened by the factual part.

Photo of part of a book page reading:
“What kind of science was this, Thoreau asked, 'which enriches the understanding, but robs the imagination'? This was what Humboldt had written about in Cosmos. Nature, Humboldt explained, had to be described with scientific accuracy but without being 'deprived thereby of the vivifying breath of imagination'. Knowledge did not 'chill the feelings' because the senses and the intellect were connected. More than any other, Thoreau followed Humboldt's belief in the 'deeply-seated bond' that united knowledge and poetry. Humboldt allowed Thoreau to weave”

Photo of the following book page, reading:
“together science and imagination, the particular and the whole, the factual with the wonderful.
Thoreau continued to search for this balance. Over the years, the struggle became less intense, but he remained worried. One evening, for example, when he had spent a day at a river, scribbling page after page of notes on botany and wildlife, he finished the entry with the sentence: 'Every poet has trembled on the verge of science? But as he plunged into Humboldt's writing, Thoreau slowly lost his fear. Cosmos taught him that the collection of individual observations created a portrait of nature as a whole, in which each detail was like a thread in the tapestry of the natural world. Just as Humboldt had found harmony in diversity, so too did Thoreau. Detail led to the unified whole or, as Thoreau put it, 'a true account of the actual is the rarest poetry?”

220 2

Anne Scheel @annemscheel.bsky.social · 31/10/2024

This is how we explained it in our primer for psychologists doi.org/10.1177/2515...

2162 16

Anne Scheel @annemscheel.bsky.social · 27/10/2024

Haven’t read the article but I’m a fan of this research line. Very thoughtfully and carefully done, IMO exemplary for how psych research should proceed. They have a nice webpage kamamutalab.org We used it as an example in this paper journals.sagepub.com/doi/full/10....

130 3

Anne Scheel @annemscheel.bsky.social · 10/10/2024

Got a print of Alexander von Humboldt’s Geography of Equinoxial Plants from my partner (he’s reading @andreawulf.bsky.social’s Invention of Nature and caused a massive Humboldt craze in our household). It’s so beautiful! Look at all those details! It’s actually all legible (if you can read German..)

1110 11

Anne Scheel @annemscheel.bsky.social · 20/09/2024

Will, I think the reason for why you may have sensed hostility from parts of the OS community is how you treated your PhD student. docs.google.com/document/d/1...

0104 10

Anne Scheel @annemscheel.bsky.social · 31/08/2024

Another* explanation, of course, is that Haidt’s target audience are not other researchers but the general public. I’ve written about why I think that’s bad for science in this blog: www.the100.ci/2021/07/08/a... * additional, not necessarily alternative

1203 20

Anne Scheel @annemscheel.bsky.social · 31/08/2024

I know statistics is hard! I don’t expect everyone to be a stats genius. Just that after studying psych, doing research for years and passing the tenure bottleneck, you _must_ have learned that you can’t just assume it’s easy and do vibes-based analyses without asking an expert

A picture of Upton Sinclair with his famous quote “It is difficult to get a man to understand something, when his salary depends upon his not understanding it!”; mirrored horizontally with an adapted quote by “Downton Sinclair”, saying “It is easy to get a man to understand something, when his salary depends upon his understanding it!”

493 9

Anne Scheel @annemscheel.bsky.social · 30/08/2024

Somehow I missed the part where Stapel plagiarises James Joyce in the book he wrote about his fraud www.psychologicalscience.org/observer/der...

1112 11

Anne Scheel @annemscheel.bsky.social · 09/07/2024

Me after learning about collider bias

2356 35

Anne Scheel @annemscheel.bsky.social · 26/06/2024

Table 1 has the info you’re looking for. 41/71 RRs were replications, 39% positive results, whereas the 30 original RRs had 50%. So it may well make a difference, but it’s nowhere near non-RRs. I like to think that RRs select for 1) questions where the answer is interesting either way and >

130 3

Anne Scheel @annemscheel.bsky.social · 29/02/2024

In his keynote at the Perspectives on Scientific Error workshop, @ianhussey.bsky.social finally answers the question who can help us carry the burden of detecting all the countless scientific errors out there: A clone-army of Nick Browns

1227 22

Anne Scheel @annemscheel.bsky.social · 18/02/2024

So you’re worried that people will 1) write papers with deliberate errors in them, 2) get them published, 3) team up with a reviewer to split the bounty with in order to 4) cash CHF500 minus the reviewer’s share, or up to CHF2,500 to get their own paper retracted?

110 1

Anne Scheel @annemscheel.bsky.social · 18/02/2024

Ad-hoc questionnaires are psychology’s untested antibodies

280 8

Anne Scheel @annemscheel.bsky.social · 06/02/2024

Very interesting and encouraging pilot at the University of Sussex — hope this can spread to other departments and universities! h/t @deevybee.bsky.social www.ukrn.org/2024/01/30/p...

05830 58

Anne Scheel @annemscheel.bsky.social · 29/12/2023

Scientists gaming their impact on policy in 3, 2, …

0173 17

Anne Scheel @annemscheel.bsky.social · 03/11/2023

Love this 10-year-old quote by Brian Haig. The hidden-moderator debate seems to have mostly faded out, but IMO this is a great counter argument to those who used to say that we should focus on the methodological differences that allegedly cause replication failures.

1134 13

Anne Scheel @annemscheel.bsky.social · 27/10/2023

It’s a most exciting day — we’re officially launching the Dutch Reproducibility Network (NLRN) with a symposium in Amsterdam! Our social media account is still on the dark side, if you want to follow along: twitter.com/nlrepro/stat...

1296 29

Anne Scheel @annemscheel.bsky.social · 01/10/2023

Haven't tried out JATSdecoder myself yet but it does sound quite powerful

010 1

Anne Scheel @annemscheel.bsky.social · 01/10/2023

Also contains super interesting p-value stats. Would love to see the full p-value distribution for each journal!

120 2

Anne Scheel @annemscheel.bsky.social · 01/10/2023

E.g., here's how median (and .75 quantile) sample sizes changed before vs after 2015 across 12 journals. Also really interesting as an index for sample-size differences between subfields:

140 4