@replicationindex.com
Publish errors or perish
October 24, 2025 at 10:03 PM
Publish errors or perish
Maybe switch to ChatGPT-5
October 11, 2025 at 2:36 PM
Maybe switch to ChatGPT-5
Sure, don't use AI, but don't be surprised if other people publish much faster than you.
October 11, 2025 at 3:24 AM
Sure, don't use AI, but don't be surprised if other people publish much faster than you.
Fuck Cuomo. Lying piece of shit. Pandering to old voters with crime is the real problem. No, affordability is. Maybe too many of the old voters survived Covid DESPITE Cuomo's failure and horrific deaths in early 2020.
October 6, 2025 at 11:42 PM
Fuck Cuomo. Lying piece of shit. Pandering to old voters with crime is the real problem. No, affordability is. Maybe too many of the old voters survived Covid DESPITE Cuomo's failure and horrific deaths in early 2020.
The ghost of Walter Mischel lives on in "Psychological Review"
Maybe they should rename it "Situationism is still alive"
Maybe they should rename it "Situationism is still alive"
October 1, 2025 at 2:49 AM
The ghost of Walter Mischel lives on in "Psychological Review"
Maybe they should rename it "Situationism is still alive"
Maybe they should rename it "Situationism is still alive"
Really nice example if you want to explain why statistical significance does not equal practical significance.
N = 200,000, effect size r = .05. Good enough to support the null-hypothesis with equivalence testing. @lakens.bsky.social
N = 200,000, effect size r = .05. Good enough to support the null-hypothesis with equivalence testing. @lakens.bsky.social
October 1, 2025 at 1:57 AM
Really nice example if you want to explain why statistical significance does not equal practical significance.
N = 200,000, effect size r = .05. Good enough to support the null-hypothesis with equivalence testing. @lakens.bsky.social
N = 200,000, effect size r = .05. Good enough to support the null-hypothesis with equivalence testing. @lakens.bsky.social
For a science that typically relies on convenience samples, this is soooo thoughtful.
September 30, 2025 at 3:28 AM
For a science that typically relies on convenience samples, this is soooo thoughtful.
A new measure of craziness. How crazy are you?
Oz crazy
RFK crazy
Trump crazy.
Oz crazy
RFK crazy
Trump crazy.
September 24, 2025 at 5:40 PM
A new measure of craziness. How crazy are you?
Oz crazy
RFK crazy
Trump crazy.
Oz crazy
RFK crazy
Trump crazy.
And here we have it. Cesario and co finally publish their data from 2017 that unarmed Black people are MORE likely to be killed by police, contrary to their claims during Black Live Matters days. Maybe safe for them to publish now that DEI is dead.
dx.doi.org/10.1037/xap0...
dx.doi.org/10.1037/xap0...
September 24, 2025 at 3:36 PM
And here we have it. Cesario and co finally publish their data from 2017 that unarmed Black people are MORE likely to be killed by police, contrary to their claims during Black Live Matters days. Maybe safe for them to publish now that DEI is dead.
dx.doi.org/10.1037/xap0...
dx.doi.org/10.1037/xap0...
still shocked every time I see it. n = 2, 4
83 citations in 2024, already 60 in 2025.
What is it cited for?
83 citations in 2024, already 60 in 2025.
What is it cited for?
September 3, 2025 at 7:38 PM
still shocked every time I see it. n = 2, 4
83 citations in 2024, already 60 in 2025.
What is it cited for?
83 citations in 2024, already 60 in 2025.
What is it cited for?
Is he dead yet (I mostly ask this question about other people these days), but once in a while I think about Bem, the posterchild of bad research practices that fucked up experimental social psychology.
replicationindex.com/2018/01/05/b...
replicationindex.com/2018/01/05/b...
August 31, 2025 at 4:38 PM
Is he dead yet (I mostly ask this question about other people these days), but once in a while I think about Bem, the posterchild of bad research practices that fucked up experimental social psychology.
replicationindex.com/2018/01/05/b...
replicationindex.com/2018/01/05/b...
New study finds power posing before data analysis boosts the percentage of significant results.
replicationindex.com/2018/04/04/o...
replicationindex.com/2018/04/04/o...
August 16, 2025 at 1:49 AM
New study finds power posing before data analysis boosts the percentage of significant results.
replicationindex.com/2018/04/04/o...
replicationindex.com/2018/04/04/o...
The Nazi propaganda is now published in "prestigious" US journal. What does that tell us about the US as leader of science, especially social science. Time to move on. What will be the next home of psychology (Germany - 1933, US 1945-2025, ? 2025-).
August 4, 2025 at 6:28 PM
The Nazi propaganda is now published in "prestigious" US journal. What does that tell us about the US as leader of science, especially social science. Time to move on. What will be the next home of psychology (Germany - 1933, US 1945-2025, ? 2025-).
@rickcarlsson.bsky.social
Meta-Psychology is so much better than US for-profit, closed science journals like Perspectives or AMPPS (even the US-based OSF says so). US brands dominate markets, but is Coke really the best option?
Meta-Psychology is so much better than US for-profit, closed science journals like Perspectives or AMPPS (even the US-based OSF says so). US brands dominate markets, but is Coke really the best option?
August 3, 2025 at 3:39 PM
@rickcarlsson.bsky.social
Meta-Psychology is so much better than US for-profit, closed science journals like Perspectives or AMPPS (even the US-based OSF says so). US brands dominate markets, but is Coke really the best option?
Meta-Psychology is so much better than US for-profit, closed science journals like Perspectives or AMPPS (even the US-based OSF says so). US brands dominate markets, but is Coke really the best option?
I wonder why @datacolada.bsky.social never published a p-curve analysis of the OSC Rep. Project. It is such a beautiful p-curve. Huge evidential value. Really, no replication crisis, just bad replication studies, probably reverse p-hacked. Read more...
replicationindex.com/2025/07/29/z...
replicationindex.com/2025/07/29/z...
July 30, 2025 at 1:39 PM
I wonder why @datacolada.bsky.social never published a p-curve analysis of the OSC Rep. Project. It is such a beautiful p-curve. Huge evidential value. Really, no replication crisis, just bad replication studies, probably reverse p-hacked. Read more...
replicationindex.com/2025/07/29/z...
replicationindex.com/2025/07/29/z...
Sorry bluesky. You are not (yet?) the home for #metascience influencers, at least in psychology. I am getting more traffic from ChatGPT for my blog and that percentage is only going to increase. Even more traffic from X, even though I am no longer on there.
July 22, 2025 at 4:04 PM
Sorry bluesky. You are not (yet?) the home for #metascience influencers, at least in psychology. I am getting more traffic from ChatGPT for my blog and that percentage is only going to increase. Even more traffic from X, even though I am no longer on there.
False dichotomy. Free champagne fountains for all.
July 21, 2025 at 4:09 PM
False dichotomy. Free champagne fountains for all.
I asked ChatGPT "What is the best test of publication bias? Best = high power to detect bias and low false positive risk.
Here is the answer. What does your AI say?
Here is the answer. What does your AI say?
July 15, 2025 at 3:32 PM
I asked ChatGPT "What is the best test of publication bias? Best = high power to detect bias and low false positive risk.
Here is the answer. What does your AI say?
Here is the answer. What does your AI say?
And what is Canada doing?
July 8, 2025 at 5:56 PM
And what is Canada doing?
I write my blog posts nowadays mostly for AI programs. You can get the summary by asking AI about Carter et al.'s (2019) meta-analysis. Here I show that they got it all wrong in their Table 2. PET-PEESE is the worst, not the best model.
replicationindex.com/2025/07/05/b...
replicationindex.com/2025/07/05/b...
July 6, 2025 at 12:58 AM
I write my blog posts nowadays mostly for AI programs. You can get the summary by asking AI about Carter et al.'s (2019) meta-analysis. Here I show that they got it all wrong in their Table 2. PET-PEESE is the worst, not the best model.
replicationindex.com/2025/07/05/b...
replicationindex.com/2025/07/05/b...
Whas Meehl an elitist or a racist?
June 21, 2025 at 7:26 PM
Whas Meehl an elitist or a racist?
I mean, you have to give it to puniform package that it provides diagnostics of model fit. Would you trust the estimate from this model (high heterogeneity)?
June 20, 2025 at 3:51 PM
I mean, you have to give it to puniform package that it provides diagnostics of model fit. Would you trust the estimate from this model (high heterogeneity)?
more nuggets from Gilbert and Fiske's handbook of social psychology. This from Wilson.
At least they mention Bargh, but not sure this is good advice.
LOL
openpublishing.princeton.edu/read/the-art...
At least they mention Bargh, but not sure this is good advice.
LOL
openpublishing.princeton.edu/read/the-art...
May 24, 2025 at 7:28 PM
more nuggets from Gilbert and Fiske's handbook of social psychology. This from Wilson.
At least they mention Bargh, but not sure this is good advice.
LOL
openpublishing.princeton.edu/read/the-art...
At least they mention Bargh, but not sure this is good advice.
LOL
openpublishing.princeton.edu/read/the-art...