I updated my post on equivalence tests using {marginaleffects} this morning.
If you want to make an argument *for* no effect (or as I like to say "for a negligible effect"), then you should be using an equivalence test.
#polisky #stats #metasci
www.carlislerainey.com/blog/2023-08...
0 replies
15 reposts
36 likes
Reposted by Jay Patel
Appreciating the centering of feminist โwho questionsโ in DโIgnazio and Kleinโs Data Feminism:
๐ธBy whom?
๐ธFor whom?
๐ธWho benefits?
๐ธWho is harmed?
๐ธWho speaks?
๐ธWho is silenced?
0 replies
2 reposts
2 likes
PSA: If you want to save a thread from a social media site fully, faithfully, and for free, use the Zotero web extension in a Chrome-based browser.
Beats the paid tools every time.
#research #phdchat
chromewebstore.google.com/detail/zotero-connector/ekhagklcjbdpajgpjgmbionohlpdbjgc?pli=1
0 replies
4 reposts
13 likes
Agreed.
And what if the next step for pre-registrations is to guide scholars to make rigorous (justifiable) choices to not only document, but also decide well?
Then it would become something like an experimental design tool.
0 replies
1 reposts
2 likes
๐ A thought experiment for researchers:
If you could click one button to share your (non-sensitive) notes from your articles in Zotero, Mendeley, PaperPile, etc. would you open those data up to the world?
#metascience
0 replies
0 reposts
2 likes
I'm incredibly excited as a sometimes-UXR to get to the bottom of how to use UXR to promote rigorous scholarly practices. Let's spark that community!
1 replies
0 reposts
3 likes
However the authors used N = 50 in the study. What should the Ns have been? I've never been alerted to this error before your posts.
1 replies
0 reposts
2 likes
Uh oh, that reminds me of a paper I read a while back. It used G*Power to compute Ns for an experiment on fear extinction.
doi.org/10.1038/s413...
The details are in this paragraph (2 groups, 2 measurement time points: pre and post).
I ran the power analysis selecting SPSS in G*Power (N = 194).
1 replies
1 reposts
3 likes
cc @mikemorrison.bsky.social for UX issues in statistical software (seems like your jam... and mine.)
1 replies
0 reposts
1 likes
Do you have a GDoc link or something like that for copy-pasting?
0 replies
0 reposts
0 likes
Reposted by Jay Patel
โGlobally, cold deaths are 9 times higher than heat-related ones. In no region is this ratio less than 3, and in many, itโs over 10 times higher. Cold is more deadly than heat, even in the hottest parts of the world.โ
Mindblowing from Our World in Data
ourworldindata.org/part-one-how...
3 replies
8 reposts
9 likes
Reposted by Jay Patel
NISO announces publication of its Communication of Retractions, Removals, & Expressions of Concern (CREC) Recommended Practice. www.niso.org/press-releas...
#retractions #ResearchIntegrity
0 replies
2 reposts
2 likes
Reposted by Jay Patel
Just a reminder that an extremely low p-value is likely if the effect is strong or the sample size is much larger than needed. P < 1e-10 isn't indicative of any wrongdoing. #stats
4 replies
3 reposts
16 likes
Reminds me of Wittgenstein for some reason:
"a wheel that can be turned though nothing else moves with it is not part of the mechanismโ(ยง271)."
0 replies
0 reposts
1 likes
Is library(stargazer) no longer maintained?
1 replies
0 reposts
1 likes
Figma offering slide design tools is a great change of pace to shake up the PowerPoint-Keynote domination:
www.figma.com/slides/
Worth trying for your next talk?
0 replies
1 reposts
1 likes
I think it might be an analog of HARKing: Hypothesizing after Results are Known.
So maybe we could call the Questionable Research Practice "Systematic Searching After Results are Known" or something catchier.
1 replies
1 reposts
6 likes
Reposted by Jay Patel
The Unjournal is looking for impactful papers to evaluate. We welcome the submission of papers by authors, and paper recommendations in the quantitative social sciences field.
To suggest research: coda.io/form/Suggest...
To submit your research: coda.io/form/Author-...
0 replies
1 reposts
2 likes
For me, the issue is the positionality statements don't attempt to correct for possible biases or suggest corrective strategies. I like them and plan to use them, but would want to supplement positionality with strategies that help others think about what I've done to avoid confirmation bias.
1 replies
0 reposts
1 likes
๐ Noise-canceling tech, passive and active, is crazily underfunded while being more important than flashy whiz-bang gizmos.
Solving this tech puzzle will noticeably improve life.
0 replies
0 reposts
0 likes
Reposted by Jay Patel
The replication crisis is often described as a crisis of false positives, but is it?
In our (re)new(ed) preprint, we show that false positives do not explain the replication crisis, and that varying rep. rates are often explained by replication sample size. #metascience
osf.io/preprints/so...
5 replies
41 reposts
94 likes
Reposted by Jay Patel
Me thinking about conferences: "They need an Introverts room! You go in and even if you know someone you can't talk to them. You just sit and chill."
Today I learned some conferences have that! They just call it a Quiet/Low Stimulation Room. Love it.
0 replies
3 reposts
9 likes
๐ Checking alignment between the Abstract and body text is key for #peerreview.
It should be part of any review/appraisal checklist.
0 replies
0 reposts
0 likes
Agreed, I'm reminded of medical articles that advocate for exactly this. I'm on board!
1 replies
1 reposts
1 likes
This is a great initiative for ECRs like Ph.D. students looking to learn about peer review. If you're already keeping up with meta-research and/or biology research, please review with us.
COI: I am a co-lead for the meta-research club.
1 replies
3 reposts
1 likes
Just import your Twitter network using Sky Follower Bridge:
chromewebstore.google.com/detail/sky-f...
1 replies
0 reposts
0 likes
EQUATOR Network is a great platform for finding reporting guidelines/checklists, but the reports are in Word and PDF.
Could we at least see them in .csv form for analysis? Imagine a metascientific study showing the overlap between criteria and mapping the landscape out.
#metascience
0 replies
0 reposts
1 likes
So... no "sensor fusion"?
0 replies
0 reposts
0 likes
Total IQ: low 40s?
1 replies
0 reposts
0 likes
Reposted by Jay Patel
New role for a Project Coordinator at COS. Perfect for an early career researcher passionate about open scholarship and looking to gain research experience, whether to then go to grad school or stay in industry.
jobs.lever.co/cos/ce551e5d...
0 replies
36 reposts
26 likes
Reposted by Jay Patel
๐งช Hey #ScienceBluesky! Would you join an open science social media mirroring service? It would ensure that all your research-related posts were citeable & open access, regardless of source platform (X, Masto, Bsky,)!
๐ RT if youโre interested!
For more context see -> osf.io/preprints/me...
1 replies
5 reposts
5 likes
Reposted by Jay Patel
The online portion of #SIPS2024 will start in less than 2 days! ๐ Join us on June 19-20 on Zoom Events for brand new hackathons, unconferences, workshops, roundtables, and pre-study posters, as well as continuations of sessions started in person! #PsychSciSky
0 replies
3 reposts
8 likes
๐ Open peer reviews are long and awkward to read. Has anyone made a habit of reading them or are they the new supplementary materials?
๐ It might help to summarize them in a visual form if the goal is to draw attention to them.
#metascience
0 replies
0 reposts
0 likes
Reposted by Jay Patel
Check out our first offering in the new series, "Breakout room," in which four researchers debate non-testable theories in neuroscience. Moderated by @neurograce.bsky.social and featuring @stephaniepalmer.bsky.social @cianodonnell.bsky.social, @dlbarack.bsky.social.
bit.ly/4cgtMsH
0 replies
1 reposts
4 likes
Do you want any economic, social science, or policy reason evaluated in a novel, journal-independent way?
Send your ideas to @unjournal.bsky.social ๐
0 replies
0 reposts
1 likes
Reposted by Jay Patel
Suggest impactful research for The #unjournal (unjournal.org) to commission for evaluation:
2 replies
1 reposts
2 likes
I've never experienced these syndromes, but I'm reminded of the cuteness overload phenomenon:
www.npr.org/sections/hea...
Maybe there's a connection somewhere such that too much of a good thing (cuteness, prettiness) causes breakdown of some kind.
0 replies
0 reposts
2 likes
"It is tempting to seek a simple single principle with which we can grasp the essence of science and toss out all else. Unfortunately, science is too complicated for that."
-- Robert Crease
0 replies
0 reposts
2 likes
IMO, falsifiability is too strict a criterion for theorizing. It can be a starting point, but modern philosophy shows otherwise.
Robert Crease is a good starting point: iopscience.iop.org/article/10.1...
1 replies
0 reposts
1 likes
Adapted from this article:
www.ncbi.nlm.nih.gov/pmc/articles...
So that explain the y-axis truncation and mislabeling. Original plot is similar:
See part B for tension headache sex differences:
0 replies
0 reposts
0 likes