Here's a new paper from the lab (w/1st author
@PaigeAmormino
and stats expert
@mercierbrett
) where we investigate how people judge the praiseworthiness of helping.
This is pretty wild: University of Washington's Department of Psychology has a 2-year TT hiring ban because they used race as the deciding factor in a recent hire (violating university policy and state law), then tried to cover it up.
Data Colada have collectively volunteered thousands of hours to improve science. Unfortunately, sometimes when you scrutinize the work of powerful people...they sue you. Together, we can stand up for science and help them defend themselves in court.
I've seen some misinformation floating around about the UCLA grad student letter. To be clear, I didn't post it on Twitter and don't support doing so with the signatories' names attached, because I don't want anyone to be doxxed or harassed.
Noticed this in the discussion of the GA bill as well (which, incidentally, was co-authored by equal numbers of men and women). The left's focus on identity might be why liberals are seeing a gender divide where none exists.
As part of the next
@fourbeerspod
episode, I took a look at a recent field study of women at nightclubs. The key finding is that women's self-objectification moderates the relationship between their clothing and self-reports of feeling cold.
@NAChristakis
@lessig
I just don’t think that happened here! Data Colada researchers approached Harvard privately. Harvard took over a year to investigate, brought in outside experts, and wrote a 1,200 page report. What part of that is caving to outrage?
Here's a new paper from the lab in which we find that regardless of party, U.S. politicians use more moral language (on Twitter and in House/Senate speeches) when their political power is lower.
Here's a new review with
@peez
(forthcoming in Advances in Experimental Social Psychology) in which we attempt to make sense of the effects of disgust on social judgments. This was way more work than we expected but I'm happy with how it turned out.
I missed 🔥🔥🔥 week, but:
1. R is a terrible programming language: fragile, limited, slow
2. R is annoying to use interactively: verbose and inconsistent with inscrutable error reporting
3. The 1 reason to use R is everyone else uses it, and there's a package for everything
New episode! "library(blackgoatpod)". More and more psychology researchers are using R. To many users it is more than just a way to do statistics - it's part of a larger approach to doing and thinking about quantitative research
New GMO paper that I'm involved with somehow: "Extreme opponents of genetically modified foods know the least but think they know the most"
Full-text here:
As psych grad programs
#GRExit
, what advice would you offer for folks submitting applications
#withoutGRE
.
@UMNPsych
, I have some ideas. What are yours?
“All else being equal, an unvaccinated 66-year old is about 30 times more likely to die, given a confirmed case, than an unvaccinated 36-year-old, and someone over 85 is over 10,000 times more at risk of dying than a child under 10.”
Having now read the Nature Communications paper that is causing all the uproar, I have two reactions. First, the causal language and policy recommendations (even hedged) seem premature given the evidence. As an editor I would have asked for changes. >>
You know, it would be nice if this thoughtful response from the authors spread as widely as the initial angry yelling about the paper. I think prominent folks who amplified misinformation have some self-scrutiny to do here.
Thank you to everyone who has voiced their concerns with our recent PNAS paper on ML models of first impressions. Conversations around ethical applications of face-processing technologies are essential. We're listening. Here are our reflections:
Hello! Are you a prospective PhD student interested in morality, politics, and/or technology? Consider applying to my lab at the University of Toronto for the 2022 academic year.
To get an idea of what we do, check out some recent papers here:
Incidentally, I am very supportive of graduate student feedback in the hiring process! In this case, of course, I think they were wrong on the merits, and also wrong to post their letter on a public website (rather than sharing confidentially with the hiring committee).
Seeing the conversation unfold on this has been frustrating, with a lot of, IMO, bad faith critiques. That most of the signatories are famous enough to be hard to fire is sort of the point. They are the ones in a position to write and sign such a letter.
A statement signed by 150 people incl. Bill T. Jones, Wynton Marsalis, Jennifer Finney Boylan, Noam Chomsky, J.K. Rowling, Margaret Atwood, and Salman Rushdie expresses concern over the illiberal trend intensified by our national reckoning.
Dear Psych Twitter: I teach an intro-level social psychology course using the Gilovich et al. textbook. It is fine, but stupidly expensive. Anyone using free social psychology textbooks they like? Feel free to RT!
right except this is also performative (aimed at weird centrist tribes who think they've transcended natural human signaling). There's only one pure unalloyed non-performative entity on twitter and that's me
New Data Colada post evaluating the claims in a recent paper in Nature. There was a preregistration (great!). But you have to *read the preregistration* to evaluate the credibility of the results.
I used to teach this finding to my MBA students. It has more than 400 citations on Google Scholar. It's about dishonesty. Turns out it's fraudulent. "Evidence of Fraud in an Influential Field Experiment About Dishonesty"
So, retraction would be transparently political, b/c people are morally offended by the findings/interpretation. Talk of "harm" is just repackaged moral outrage. Where we see harm depends on our moral/political outlook. Enshrining "harm" as a reason to retract just bakes in bias.
@jacasiegel
@Sakiera_Hudson
Stapel, Sanna, Smeesters, Forster, Wansink, Ariely. All men. I was 1 of 3 ECRs who blew the whistle on Stapel. I had to talk to my wife about what would happen if I lost my job. Takes like this are why I chose to remain anonymous until tenure.
First Episode! In Search of the Campus Free Speech Crisis | We tackle the alleged free speech crisis on campus. Is there reason to worry or are reports of left-wing intolerance overblown? Also: why are we doing this?
Subsribe to our show!
But, also, calls to retract, some from well-known researchers, are way off base. If we are retracting papers for imperfect measures, over-enthusiastic causal claims, and some premature policy recommendations, we are retracting many many many papers.
Curious about how the non-moral becomes moral over time? Check out our new paper in JPSP on how meat consumption becomes moralized. W/Matt Feinberg, Chloe Kovacheff, & Rimma Teper (
@rimkat
), who all did more than I did. [full-text: ]
How did the letter get out? One of the grad students posted it to Google Drive publicly (i.e., accessible to anyone in the world) on January 25, 2023. A link to the email was then sent to many (100+) members of the UCLA psych community (see screenshot).
Here's a new paper from the lab in which we find that when politicians (U.S. presidential candidates and members of Congress) use more moral language on Twitter, they are retweeted more.
@jenn_rubin
At that price it is not affordable for anyone who is not rich. That's more than 70K a year. The median US household income is less than that (~67K).
@DrJessTaylor
@ashleyruba_phd
These figures are completely wrong. RELX (Elsevier's parent company) includes multiple divisions (Lexis-Nexis, conventions, etc.). In 2022 their *revenue* was ~10.5B, operating income ("profit") was ~$2B. For comparison: Apple ~$114B, Alphabet (Google) ~$60B, Amazon ~$12B.
I hadn't seen this before: part of the forensic report on Gino's papers that was commissioned by Harvard. It only covers the PNAS sign-at-the-top paper but it looks incredibly damning:
At some point we have to decide whether we want to know what's true, or whether we'd rather make progressives comfortable. I would rather know what's true.
Here in Canada, the Liberal party just won a plurality of seats (and will stay in power) despite the Conservative party winning the most votes. Haven't seen much about how this is unfair, undermines democratic legitimacy, etc. Do Canadians not care or is it my (liberal) bubble?
You'll probably have seen tweets urging you to read this paper. But seriously, read this paper. It's the most I've enjoyed reading a paper in, maybe, ever.
If (like me) you've always found logistic regression to be a little mystifying, this is the episode for you. Listen and you'll never have to google "how to interpret odds ratio" again!
It’s Quanti-TUDEsday!
S3E17: Logistic Regression: 2 Logit 2 Quit
In today's episode, Greg & Patrick explore the generalized linear model as a powerful framework for building regression models for binary and other discrete dependent variables.
Canadian friends: you might be tempted to try
@FlairAirlines
because they can be much cheaper than Air Canada or Westjet. My advice for you: if you would like to get where you're going, do not do this.
I have an article out today about police violence. In a comprehensive assessment of the data available, I show that police killings are rising and it's basically because the police have been engaged in a 20-year-long riot against the republic.
Finally, I know this analysis comes in the context of a fierce debate about the parameters of acceptable criticism online. My feeling is that when a paper is published it is open to critique. Distributed organized skepticism is the business we are all in.
@BrianNosek
I chimed in over there as well. I think it's good for her followers on LinkedIn (largely non-academic) to hear what researchers think of this.
So, the results in this paper are dependent on the specific analytic choices made. There is only one path through the garden of forking paths that produces p < .05. After doing these analyses, I have less confidence in the results. Some lessons follow:
Lesson 3: I was only able to do this analysis because the data and materials were open. Open data are great. But (to use a
@siminevazire
analogy) transparency only opens the hood. You still need to see what the engine looks like.
@wgervais
@dingding_peng
@BrianNosek
@LauraMKoenig
I get your general point but in this case I think it's outweighed by the value of seeing what people are willing to write anonymously. This excerpt is unambiguously bad and I doubt whoever wrote it would be willing to stand behind it publicly...but yet it can block publication
You know, maybe the lack of a control group should be mentioned before paragraph 21.
Many Post-Covid Patients Are Experiencing New Medical Problems, Study Finds
one of the professors in our department was on a podcast and said he felt "energized" by the "jubilation" of the hamas attack and you know how it is, espresso for some terrorism for others
This might be my favorite episode yet. Simine and Robb have a discussion in which they take each other's views seriously but aren't afraid to disagree. I was so happy to be able to co-host this.
P.S. Sorry for drinking bourbon instead of beer.
New episode!
The COVID debate (with Robb Willer & Simine Vazire)
@RobbWiller
and
@siminevazire
join the podcast to debate whether social science, in its current state, can usefully contribute to the COVID19 response. Bonus: Guess who didn't drink beer?
@Atul_Gawande
@PardisSabeti
This is a bad, terribly biased take. Much of social psych is quietly moving forward with better research practices. The blow ups are generally when researchers with “threatened” findings refuse to accept that they might be wrong.
Lesson 1: Preregistration would have been useful here. It is hard to know why particular paths were chosen without a preregistered rationale.
Lesson 2: High p-values are an imperfect heuristic for fragile results, but a reasonable one. High p-values *should* raise eyebrows.
Let’s settle this, b/c I personally think this is one of the dumber arguments of this decade (philosophy non-polite etiquette enabled) - On the claim that we need more political conservatives to do better science, akin to having more women and minorities.
Most important takeaway: it is trivial to get around MTurk location restrictions. You have to assume MTurk workers might be anywhere in the world and might not speak English well or at all.
MTurk Bot-gate 2018, Update:
Apparently the mTurk bots are humans, NOT bots. They are shady mTurkers who use server farms to conceal their identity and crack through surveys. Sean Dennis's working paper below spells out the problem--and how to fix it.
I'm re-watching Star Trek:TNG and...why is everyone so weirdly blasé about Q? He has literally godlike powers to control matter and time and the crew of the Enterprise mainly seems just kind of annoyed he’s bugging them. Q should be causing a crisis in 2300's theoretical physics!
Careful, detailed write-up by
@FiveThirtyEight
's
@cragcrest
of a just-accepted paper trying to empirically assess political bias in social psych. With many wonderful co-authors, none of whom seem to be on Twitter.
I'm excited to announce the 18th annual JDM preconference at SPSP, in-person in Atlanta February 23, 2023! We have a stellar line-up of speakers and submissions for the grad student/postdoc poster session are open (deadline November 15).
@datingdecisions
@fourbeerspod
In my lab the grad student is the lead author on most papers. I think that is common. So that would mean you can't critique many papers. This does not seem desirable to me.
@minzlicht
@jinxgoh
Full disclosure: we've had him on our podcast (
@fourbeerspod
) and he seems smart, fair-minded, and reasonable. I've read a lot of attacks, not a lot of substantiation.
@Sakiera_Hudson
I'm sorry, you're probably just kidding around, but I think you should re-think this framing. People who detect fraud are doing a vital service as uncompensated volunteers. And the DC folks have been more than careful. "Witch hunt" is a ridiculous description here.
New episode! I love how you hurt me (with Paul Bloom)
Yoel and Mickey welcome
@paulbloomatyale
to the show to discuss the potential benefits of pain. Why do we sometimes choose to suffer? Are there benefits (to self or society) to being a painful person?
I appreciate this long and candid discussion of what happened to me at UCLA from an insider. It's worth reading it all but it seems very clear that what motivated the faculty decision was an ideological objection to a years-old podcast about diversity statements.
In January, Dr. Yoel Inbar (
@yorl
) interviewed for but did not receive a job offer from UCLA’s Psychology Department. This case was not handled properly. See my take in the attached blog. Please retweet (re-X?) if this concerns you.
I'd be very interested in the strong argument against this kind of research (ideally including readings). I see a lot of moral outrage in the replies/quote tweets but I would like to know more about why. As an outsider it seems sort of baffling.
@NAChristakis
@StuartJRitchie
@lessig
Yeah, I agree that’s bad! It just doesn’t apply here. There was no public pressure because no one knew until Harvard had made their decision (which was because DC kept it quiet until then).
Back in the U.S., staying in a lovely New England town where outdoor masking is required at the same time that indoor dining is open (and packed!) 🤦♂️
I know little about Angus Deaton but the second part of this is a master class in how to be unpersuasive. Galef is, as usual, smart and thoughtful. Deaton is so closed-minded and condescending I wanted to throw my phone out the window. Man, the arrogance of some academics...
On the latest
@rspodcast
episode, I argue with economist & Nobel Laureate Angus Deaton about
(1) What's causing "deaths of despair" and
(2) Is effective altruism a good thing?
Our conversation got... lively :)
I may be biased, but I think this article about the (incomparable) Paul Rozin's research on disgust is awesome. It starts with an anecdote about colostomy bags and just gets better from there.
Can't stop thinking about this piece. How did we end up in a place where parents are spending the night on a NYC sidewalk in February to get their kid into the right nursery school?
To say nothing of: the claim of harm depends on knowing a lot about facts and consequences. Maybe being mentored by more women actually does decrease your future citations. If so, wouldn't we want to know, and to know why? (FTR I'm not convinced by the claim, but not the point.)
Finally: enforcing agreement through power rather than persuasion works only so long as you are in power. It is extremely fragile. I would rather persuade than enforce because persuasion is much more durable.
The moralization of food preferences. Includes GE negativity, health, and "processed = bad" heuristic, but also class/aesthetic preferences for small-scale, artisanal, traditional. Lots of interesting research still to be done here.
As someone who typically uses MLM to do the "wrong thing" (i.e. to account for clustered data when I really only care about average slopes), I found this really eye-opening. Also funnier than a podcast about quant methods has a right to be!
It’s Quanti-TUDEsday!
S2E29: Multilevel Models... The Often Unnecessary Green Monster
G & P fulfill a legal obligation to interview Dan McNeish of Arizona State University about why you probably *don’t* need multilevel models even with multilevel data.
Coming next week podcasting superstar
@yorl
joins us to talk about
@talyarkoni
's "The Generalizability Crisis". Will
@tamler
and
@peez
ever be able to talk about a social psych study again?
Hi twitter, I have a career update! Please allow me to (re)introduce myself:
My name is Andre Wang, and I am a social psychologist. This summer, I will be joining the University of Toronto (
@UofT
) as Assistant Professor of Psychology! I am excited to launch my lab at
@UTSCPsych
!
Now accepted in principle at Nature Human Behaviour: "Moral thinking across the world: Exploring the influence of personal force and intention in moral dilemma judgments." Great work team! More info on next steps coming soon 🎉
I know I've heard the argument that the effect sizes from social psychology experiments are meaningless, because experiments are unrepresentative of the real world. But now I can't find the source. Does anyone know a cite?
It’s Quanti-TUDESDAY again!!
Instead of waiting 2 weeks as planned, we’re dropping the next episode of Quantitude... right now!
Check out — Episode 2: (Statistical) Power Struggles
What happens in this episode?
Patrick and Greg grumble about...
@Aella_Girl
I think most respectable journals meet that bar. Which those are though depends on the area. One problem for outsiders is there's so many low-quality journals it's hard to tell the good from the bad. Really, you need an area expert.