Do girls outperform boys on AP Calculus tests if you don't ask for gender before the test? Social scientist Joshua Aronson presented a false "stereotype threat" effect an unknown number of times. He simply flipped the girls and boys bars in his graph, showing the opposite of the true effect. Here's my report on how the false claim spread far and wide.
Joe is mostly over at Medium, and Twitter.
I'm sorry for not keeping this page up to date. I'll have lots of detailed updates on various lines of work ASAP. The short version ("short" by Joe standards, at least):
I'm in training.
I'm building skills for the next phase of my scientific work. In general, social psychologists are poorly trained, and many of their research methods and papers are invalid. They commonly use linear models incorrectly, and their bread-and-butter method is that which requires the minimum possible calorie expenditure: run quick studies using people who are already in the building and who have no choice but to participate in their studies – college kids who have chosen to take an introductory psychology course at whichever (Western, usually American) institution the researcher happens to find himself employed, who are compelled to participate in these studies for course credit, and where said studies usually consist of having the kids sit in front of a computer, read "primes", and answer surveys...
So I'm working on new, valid research methods, and looking at how to answer more useful questions about human psychology than are typically addressed by academic social psychologists. I've frozen a few of my papers until I finish some of my methodological work/training.
Some of my training is focused on better statistical methods, like latent class modeling. I'm also fascinated by what's been happening in the private sector with deep learning, and I'm trying to figure out if there might be fruitful applications there for social science. I should note however that even simple descriptive statistics would actually be an epistemic advance over the inferential statistics that academic social psychologists typically misuse, and will often better illuminate the phenomena under study. In any case, I'll probably release an R package or two at some point.
I'm going to debunk a fair amount of bogus research.
Social psychology has turned out to be a massive disappointment. It's clear at this point that most of the claims made by academic social psychologists in recent decades are simply false. I wasn't prepared for this at all. Accepting and adapting to this reality has been uncomfortable, and I've struggled to come up with a good strategy for dealing with this situation over the next ten years or so. One of the challenges is that I don't want to be known primarily as a debunker, although I think social science would benefit enormously if we had a few dozen full-time debunkers right now. I want to focus mostly on my own substantive research.
The flip side is that I think debunking false claims – and getting the media to widely report that the claims were false – might be just as beneficial for our civilization as even the most exciting original research. I think academic social psychology has substantially harmed our civilization by circulating so many false claims about human psychology and human nature. In any case, I'm trying to implement a long-term strategy where most of my work is original research on topics like envy, positive psychology, self-esteem, narcissism, the psychology of environmentalism and the ways in which it may function as a religion, accurate climate science consensus estimates, etc.
Medium, Weebly, and platforms
Most everything I've written lately is at Medium. I should have made this clear a while ago.
This website is hosted by Weebly, which also provides the back-end blogging system (kind of like WordPress). As it happens, Weebly is terrible, and I've been so frustrated with it that it affected my motivation to write/post.
To give one example, this here post is my third attempt in the last hour. In Microsoft Edge 15, Weebly wouldn't let me create links – I literally couldn't highlight any text in the post, which is the first step in creating a link. In Chrome 59, Weebly wouldn't let me login, even after I allowed third-party cookies (and I've disabled the ad-blockers for Weebly in every browser, so there's no issue there). So it's Firefox to the rescue – only in FF have I been able to both login and create links.
I promised a new, awesome website a long time ago, and haven't yet delivered. It always got bumped in favor the above-mentioned training and lots of other things. It will happen. I'm actually building two websites, which leads me to...
The Valid Science Center
I'm going to launch a 501(c)(3) – the Valid Science Center – to provide a formal channel for my work and a proper funding mechanism. (Now's a good time for you to think about how much you can contribute – this will be one the better causes you could get behind.)
It will likely start small, with just one Mexican, since I don't really need a staff to do anything I plan on doing in Phase 1. Social psychology is fairly easy, even when using valid methods – we don't need special lab equipment, we're not running PCRs or assays or anything like that. But a staff would make me more productive, and more researchers would be awesome, so I have growth in mind. (I also think it's healthier for me to work around other people, like a typical workplace, as opposed to being alone all day in a home office – I'm mostly an extrovert, and I think I've done the home office thing for too long.) I'm open to working with an existing nonprofit or think-tank, which also neatly solves the home office solitude problem.
I'm going to encourage other social scientists to start their own 501(c)(3)s or similar legal entities, so they can operate as independent scientists. I think this would be an extremely good thing for our civilization – we need clean, valid social science research that isn't crippled by the extreme political bias and methodological laziness of academic social psychology. We need scientific findings to be true most of the time. I think academic social psychology in particular is at risk of being defunded/dissolved in the near future, but if it continues to exist, it would be better for the world if it were a mere subset of social psychology as a field, instead of isomorphic with it. I'm not convinced that we can achieve intellectual/political diversity or methodological seriousness in academic social psychology anytime soon – far better would be to just step out of that muck and sprout many different homes for research out in the big, wide world, instead of prostrating ourselves before these bizarre and tiny academic tribes. Scientists should cultivate cognitive independence and integrity, and a diverse range of scientific settings would certainly help with that.
In short, we need more Ethan Perlsteins. What he's done is somewhat harder than the task before me. He's my hero.
I'll probably end up in academia eventually, the Valid Science Center in tow, but I'd prefer to go in with tenure, after I've done a lot of good research. Academia in the US is a spooky and intolerant place, not well-suited for serious scholars. A mundane non-leftist like me could be strung up before a Northwestern-style kangaroo court every other Tuesday for speaking or writing non-leftist thoughts. The scenario I favor is tenure at a university with a strong free speech code.
FYI, I'm not at all finished with the Lewandowsky and Cook fraud cases. That's still active. More updates soon...
Hey y'all... If you see weird text formatting on this site, like enlarged text sizes for specific words or phrases, that's not me — it's a Weebly technical error that just randomly enlarged portions of old content on the site.
I would never litter a post or essay with enlarged words or phrases — I'm not a crazy person. I've fixed a couple of spots I noticed (e.g. on my About page, which was so embarrassing...), and please email me at email@example.com if you notice any others.
Note: I am not talking about my haiku about Dobby. The variation in text size there is deliberate, carries substantive meaning, and I see it as a breathtaking innovation in haiku construction... (I assume it's not my innovation.)
Again, please let me know if you see anything – thanks!
I have a new essay up at Heterodox Academy discussing the claim that science is self-correcting. I also dig a bit more into what we mean by "bias".
As a reminder, Heterodox Academy is a motley crew of scholars concerned about ideological conformity and intolerance in modern academia. We're privileged to have Jon Haidt, Steven Pinker, Phil Tetlock, Judith Curry, Scott Lilienfeld, Amy Wax, Lee Jussim, Jarret Crawford, and many other outstanding scholars.
I offer some 3 am musings on the possibility.
I wrote this essay last night after reading about survival-relevant memory research (within the broader affective memory paradigm.)
Angela Chen wrote a nice article on our BBS paper over at Pacific Standard Magazine. The article digs into some specific issues that have not been discussed before in general media, like bias in how we measure openness to experience and other variables.
This is the biggest thing I've been working on this year. Read about it here.
José L. Duarte
Social Psychology, Scientific Validity, and Research Methods.