Chris Crandall
@chriscrandall.bsky.social
Social psychologist. Mediocre at so many things. Good at a few, I sure hope.
Sure.
Praxis is a vast wasteland of bad ideas.
Praxis is a vast wasteland of bad ideas.
November 20, 2025 at 3:32 PM
Sure.
Praxis is a vast wasteland of bad ideas.
Praxis is a vast wasteland of bad ideas.
Taken out of context, these ideas easily support autocratic regimes. “What is truth?” An artifice!
It’s key to build an intertidal regime which gives strong guidance for what is better science, better theory, better “belief.” Otherwise it’s RFK, Jr all the way down.
It’s key to build an intertidal regime which gives strong guidance for what is better science, better theory, better “belief.” Otherwise it’s RFK, Jr all the way down.
November 20, 2025 at 4:22 AM
Taken out of context, these ideas easily support autocratic regimes. “What is truth?” An artifice!
It’s key to build an intertidal regime which gives strong guidance for what is better science, better theory, better “belief.” Otherwise it’s RFK, Jr all the way down.
It’s key to build an intertidal regime which gives strong guidance for what is better science, better theory, better “belief.” Otherwise it’s RFK, Jr all the way down.
Thanks for all the extra info.
November 20, 2025 at 1:49 AM
Thanks for all the extra info.
Depends upon “standard” I guess. It won’t show discriminant validity, for sure.
But also, it really does predict some DVs of relationship functioning.
One person’s contaminant is another person’s predictive validity (variant if old saying).
Just need reasonable standards of validation.
But also, it really does predict some DVs of relationship functioning.
One person’s contaminant is another person’s predictive validity (variant if old saying).
Just need reasonable standards of validation.
November 19, 2025 at 6:06 PM
Depends upon “standard” I guess. It won’t show discriminant validity, for sure.
But also, it really does predict some DVs of relationship functioning.
One person’s contaminant is another person’s predictive validity (variant if old saying).
Just need reasonable standards of validation.
But also, it really does predict some DVs of relationship functioning.
One person’s contaminant is another person’s predictive validity (variant if old saying).
Just need reasonable standards of validation.
Can’t see much on my phone, sorry. Big shift recently toward white students at HKS, plus international students. Maybe not so much in this view. Thx for info.
November 19, 2025 at 6:03 PM
Can’t see much on my phone, sorry. Big shift recently toward white students at HKS, plus international students. Maybe not so much in this view. Thx for info.
Yes, very plausible.
November 19, 2025 at 5:59 PM
Yes, very plausible.
Just let the reviewers do this work; they are excellent at it.
November 19, 2025 at 2:25 PM
Just let the reviewers do this work; they are excellent at it.
But surely discriminative validity work would go a long way here and elsewhere.
November 19, 2025 at 2:24 PM
But surely discriminative validity work would go a long way here and elsewhere.
The authors don’t quite say it’s meaningless, but rather it’s contaminated by overall sentiment (valence).
The correlations that emerge are about “my relationship has very good” more than “Saturn.”
More (or less) than halo effects?
Remains to be seen.
The correlations that emerge are about “my relationship has very good” more than “Saturn.”
More (or less) than halo effects?
Remains to be seen.
November 19, 2025 at 2:23 PM
The authors don’t quite say it’s meaningless, but rather it’s contaminated by overall sentiment (valence).
The correlations that emerge are about “my relationship has very good” more than “Saturn.”
More (or less) than halo effects?
Remains to be seen.
The correlations that emerge are about “my relationship has very good” more than “Saturn.”
More (or less) than halo effects?
Remains to be seen.
He has a contract which requires it.
You want him to out? Harvard has to follow due process or pay some $$$$$$$$s.
Public opinion could make him resign. I’d sign up for that.
You want him to out? Harvard has to follow due process or pay some $$$$$$$$s.
Public opinion could make him resign. I’d sign up for that.
November 19, 2025 at 2:17 PM
He has a contract which requires it.
You want him to out? Harvard has to follow due process or pay some $$$$$$$$s.
Public opinion could make him resign. I’d sign up for that.
You want him to out? Harvard has to follow due process or pay some $$$$$$$$s.
Public opinion could make him resign. I’d sign up for that.
Have you never taught a large lecture class? Back row seats packed, front row seats sparse.
Just can’t tell from this video.
Just can’t tell from this video.
November 19, 2025 at 2:15 PM
Have you never taught a large lecture class? Back row seats packed, front row seats sparse.
Just can’t tell from this video.
Just can’t tell from this video.
Retribution without due process.
Can you imagine this being settled without truly expensive litigation, and Summers winning extra $$$s in a settlement?
Can you imagine this being settled without truly expensive litigation, and Summers winning extra $$$s in a settlement?
November 19, 2025 at 2:14 PM
Retribution without due process.
Can you imagine this being settled without truly expensive litigation, and Summers winning extra $$$s in a settlement?
Can you imagine this being settled without truly expensive litigation, and Summers winning extra $$$s in a settlement?
One way to know is if the data came back “weird” of which there are many varieties. Correlations where they should & shouldn’t be, various inconsistencies, failures of appropriate distributions & spread, etc. (Some P’s would have to be MAGA, some “antifa” and so on.)
I don’t think it’s impossible.
I don’t think it’s impossible.
November 19, 2025 at 2:12 PM
One way to know is if the data came back “weird” of which there are many varieties. Correlations where they should & shouldn’t be, various inconsistencies, failures of appropriate distributions & spread, etc. (Some P’s would have to be MAGA, some “antifa” and so on.)
I don’t think it’s impossible.
I don’t think it’s impossible.
I would be so itching to sue that person for fraud. I suppose the IRB and Research Office would be against it.
November 19, 2025 at 2:07 PM
I would be so itching to sue that person for fraud. I suppose the IRB and Research Office would be against it.
This is work for the experts who run the services. MTurk has obviously screwed the pooch; Prolific still seems OK; Verasight aggressively patrols their pool of responders.
I’m confident I can get through my career in one piece. 😉
The real lesson is to NOT rely on online data for one’s research.
I’m confident I can get through my career in one piece. 😉
The real lesson is to NOT rely on online data for one’s research.
November 19, 2025 at 2:05 PM
This is work for the experts who run the services. MTurk has obviously screwed the pooch; Prolific still seems OK; Verasight aggressively patrols their pool of responders.
I’m confident I can get through my career in one piece. 😉
The real lesson is to NOT rely on online data for one’s research.
I’m confident I can get through my career in one piece. 😉
The real lesson is to NOT rely on online data for one’s research.
How can there be any doubt?
November 16, 2025 at 3:17 PM
How can there be any doubt?
I was responding to the “hasn’t been replicated” part of that sentence. Low-quality is low quality.
November 12, 2025 at 1:18 PM
I was responding to the “hasn’t been replicated” part of that sentence. Low-quality is low quality.
I share your concern about nutrition studies, which are notoriously difficult in humans (you can randomly assign diets to rats, but . . .).
November 12, 2025 at 12:41 PM
I share your concern about nutrition studies, which are notoriously difficult in humans (you can randomly assign diets to rats, but . . .).
I want to suggest that one well-done study is better than no studies at all. Replication is undeniably useful in assuring reliability—but knowledge begins with the first study, not the second.
November 12, 2025 at 12:36 PM
I want to suggest that one well-done study is better than no studies at all. Replication is undeniably useful in assuring reliability—but knowledge begins with the first study, not the second.
There are all kinds of ways to limit these problems without the draconian, expensive, slow, and limited-in-scope methods proposed. Maybe your lab can do only RRs. So do only RRs! But why project your narrow interests on all scientists everywhere?
November 12, 2025 at 12:34 PM
There are all kinds of ways to limit these problems without the draconian, expensive, slow, and limited-in-scope methods proposed. Maybe your lab can do only RRs. So do only RRs! But why project your narrow interests on all scientists everywhere?
The biggest barrier is that there are ALL KINDS of science that don’t lend themselves to RRs. Authoritarian limitations from high prestige scientists is not the best way to speed scientific progress (see Feyerabend, 1975).
November 12, 2025 at 12:32 PM
The biggest barrier is that there are ALL KINDS of science that don’t lend themselves to RRs. Authoritarian limitations from high prestige scientists is not the best way to speed scientific progress (see Feyerabend, 1975).
Aggressively limiting the kinds of scientific questions that can be asked. No better way to shrink science than de-funding it completely (or outlawing science) than requiring only RRs.
Seriously, let a thousand flowers bloom. We can judge post-publication, too, etc.
Seriously, let a thousand flowers bloom. We can judge post-publication, too, etc.
November 12, 2025 at 12:29 PM
Aggressively limiting the kinds of scientific questions that can be asked. No better way to shrink science than de-funding it completely (or outlawing science) than requiring only RRs.
Seriously, let a thousand flowers bloom. We can judge post-publication, too, etc.
Seriously, let a thousand flowers bloom. We can judge post-publication, too, etc.
🚫
It’s a comment that is not allowed. 🚫
It’s a comment that is not allowed. 🚫
November 11, 2025 at 10:04 PM
🚫
It’s a comment that is not allowed. 🚫
It’s a comment that is not allowed. 🚫
would say that for predicting individual behavior, the IAT is a weak instrument.
By contrast, at the population level, comparing states or regions with large samples, the IAT can be pretty useful Trends over time, divergence between race, gender, and fat prejudices can be usefully tracked by IAT.
By contrast, at the population level, comparing states or regions with large samples, the IAT can be pretty useful Trends over time, divergence between race, gender, and fat prejudices can be usefully tracked by IAT.
November 7, 2025 at 4:38 AM
would say that for predicting individual behavior, the IAT is a weak instrument.
By contrast, at the population level, comparing states or regions with large samples, the IAT can be pretty useful Trends over time, divergence between race, gender, and fat prejudices can be usefully tracked by IAT.
By contrast, at the population level, comparing states or regions with large samples, the IAT can be pretty useful Trends over time, divergence between race, gender, and fat prejudices can be usefully tracked by IAT.