chriscrandall.bsky.social
Social psychologist. Mediocre at so many things. Good at a few, I sure hope.
328 posts
2,298 followers
173 following
Active Commenter
comment in response to
post
Just define “progress,” and operationalize it, measure it, and test it.
OK, not easy, but they’re not even trying.
comment in response to
post
And not even just psychology, but a fairly narrow band of so-called hypothesis-testing, looking-down-their-long-nose-at-observation-and-description style of psychology.
comment in response to
post
It gives these vibes.
comment in response to
post
A guy’s gotta hope.
comment in response to
post
Sure. But I believe the implicit goal is to elect folks around the country. Maybe control the Senste or something.
comment in response to
post
I’m reluctant to believe the NYC primary voters are representative of the entire nation.
comment in response to
post
Also, a reminder that some of the "less seasonal" choices of location are designed to keep SPSP affordable. Believe it or not, in February Chicago is cheaper than Miami.
comment in response to
post
The Duhem-Quine hypothesis, as an observation of necessity, tells us the 2 areas of progress can’t be separated.
The kinds of questions we ask depends on having methods to answer them (at least within the normal imagination).
The answer is to populate more of science with creative geniuses.
comment in response to
post
A maple tree going 75 mph would be something to see!
Glad you’re ok.
comment in response to
post
I don't know the difference between "true division of labour and expertise" and a true Scotsman, but the obvious case is when you bring in a statistical expert to do something more advanced than the other authors can manage. I assume it's common--I've done it a fair few times.
comment in response to
post
With a Ghibli engine!
comment in response to
post
This is correct.
Nevertheless, large collaborations should exist bc of division of both labor and expertise. The question is how many people should—and must—reproduce the analyses independently? N>0, certainly.
comment in response to
post
Joe: I’m reminded of the saying from the pre-web internet: “Do not feed the trolls.”
comment in response to
post
If you’re willing to do your own homework, go to Google Scholar and look at the papers that cite D-K, and you can figure out which ones are replications. That’s what scientists and scholars do.
comment in response to
post
Your nitpicking select details rather than the whole sweep of the research is classic Dunning-Kruger.
comment in response to
post
This account is a narrow & incomplete account of science publishing.
There is significant flow of funds to scientific societies from publishing (esp. when the societies own the journals) which societies in turn build science, promote students, share science.
comment in response to
post
Yikes!
comment in response to
post
“Reproducibility crisis” and half of knowledge turning over in 8 years are two different things. The first is “non-advance” and the second is “progress.”
comment in response to
post
I wonder how your research integrity training came up with that “about 50%” statistic.
comment in response to
post
Hard agree. Not only is there the focus on the wrong thing, which you cover, but also a fairly narrow understanding of all the kinds of methods, targets, procedures, and questions that people ask in science, which do no fit the prescriptions.
The paths to robust findings are many.
comment in response to
post
(became something . . .)
comment in response to
post
This documentary is 🔥. It starts in this lovely way, and then OMG, it be knew something else even better!
comment in response to
post
It’s important not to simply bc something is easy, then it must be pervasive.
Virtually every living human could shoplift or dodge their taxes—it is shockingly easy, incentivized by self-interest, and *almost* universally eschewed.
comment in response to
post
Here's an out/insider view: Went to an excellent large state university for undergrad and grad. Taught briefly at an Ivy.
The secret: Most are just upper-middle class kids who worked hard and are smart but def. not geniuses.
Effects of Ivy: (1) Trademark. (2) Networking. (3) Confidence.
comment in response to
post
For the Cohen's dz, the original/replication r = .46, for the partial eta-based effect sizes, the r = .23.
comment in response to
post
Thanks. I wanted to establish my credibility by failing to see the "DOWNLOAD PDF" button. Is it working?
comment in response to
post
These are pretty high standards for deciding on “replicated,” and these %age claims are low as a class.
For “field-wise” claims, I prefer to see a correl/scatter of 1st-time/2nd-time effect sizes. Did they have? Pay-walled.
comment in response to
post
That’s a claim that’s made. I’m unsure if any evidence beyond “it stands to reason.”
What is science without progress?
comment in response to
post
It looks useful!
(Although I think HARKing is a little too maligned . . . it's a perfectly acceptable practice in the face of replication within the same paper.)
comment in response to
post
It’s the interpretation, not the data that is suspect? I haven’t read it yet, no. I couldn’t find a link or a citation in the newspaper article.
comment in response to
post
Is it possible that concentrations of Alaska Natives are in rural areas with low candidate contact, such as Egegik, Kotzebue, or Naknek? (Yes.)
comment in response to
post
Was the study done poorly?
comment in response to
post
He used to be really credible. Will he be credible again?
comment in response to
post
But also, do not underestimate how many physics researchers are doing pure exploration. Astronomy is a good example, plasma physics is another.
I think the key advantage is shared understanding of concepts and dependent variables.
comment in response to
post
Sherman, J. W., & Rivers, A. M. (2021). There’s nothing social about social priming: Derailing the “train wreck”. Psychological Inquiry, 32(1), 1-11.
-Thanks
comment in response to
post
I like the main thrust of the argument, but as a social psychologist, I'd like to say 2 things:
1) We're the poster child in part because we addressed the issues early & vigorously
2) So-called "social priming" has not great, but social priming has almost nothing to do with social psych (cite below)
comment in response to
post
That’s right.
comment in response to
post
Bayes and Markov in one!
comment in response to
post
I’m sure that’s right.
comment in response to
post
You took my invitation to clarify the differences as an opportunity to insult.
It’s a very good example of some of the bro-pen science issues.
comment in response to
post
I think you disagree on the key point that @jbakcoleman.bsky.social & others (my hand up!) is that some of the rhetoric of the reformers (pointing at Ioannidis) has been much more helpful in support the EO than the people dragging their feet on the "good reforms" (e.g., power, open data), etc.
comment in response to
post
(use the "+" to create a thread in your replies.)
comment in response to
post
"center" as a verb in this context is too vague for me to understand your recommendations. What are you saying?
comment in response to
post
. . . demonstrations of the effect, so p-values are rarely the basis of judgement. (Or “87% of these 250 correlations are p<.01,” which is really description, not NHST.)
I think it’s important that guidelines be flexible to meet the real methods scientists use.
Data analysis serves methods.
comment in response to
post
I’m for all that. I wish this approach were more “front line” in discussions.
Most of my work isn’t much based on p-values (when I report r=.95 with N=125, I don’t report the p-value; I discuss why it was observed).
I’ve very rarely published anything “important” w/o 3 independent . . .
comment in response to
post
But probably you were taught the Duhem-Quine hypothesis, that all tests of ideas are a conjunction of the hypothesis and its operationalization? And bc of this, null effects/disconfirmation can be ambiguous, bc it’s hard to tell if it’s a “failure” of the hypothesis or of the operationalization.
comment in response to
post
Cool.
It’s the Manichean approach I’m most concerned about. Sometimes transparency is less possible (obscuring sample source, limiting other’s access to data bc participants won’t allow), etc. These are not “not science” even though they are less ideal. Sometimes, the best we can do must be OK.
comment in response to
post
I’m not against most of the reforms. I’m against the draconian approach to forcing it for all, bc it marginalizes some good science.
I like Feyerabend’s spirit in “Against Method.”