Introducing: Somewhat Unlikely
What I'm going to do in this newsletter and why you should read it.
Most interesting things in the world are ‘somewhat unlikely’.
In fact, most Interesting Things that we ever hear about in our lives are interesting because they’re somewhat unlikely (read: unusual). Yet, the headlines of new stories and peer-reviewed scientific publications make it seem otherwise; like interesting things are happening everywhere all the time.
But, if we actually zoom out and take a broader bird’s eye view of science, there’s a universal, less exciting conclusion about Interesting Things — across different types of social phenomena, different regions and time periods, and different disciplines — that comes in another related phrase: ‘it depends’.
Unfortunately, for a variety of factors (the abysmal failures of the peer review system chief among them) scientists today don’t get rewarded for saying ‘somewhat unlikely’ or ‘it depends’. This is ironic because if you actually ask most responsible social scientists to make a prediction (as journalists and soon-to-be disappointed students routinely do) about whether an Interesting Thing will or did happen — well, you can Venmo request me for $100 if they don’t say either of these phrases.
The goal of this newsletter is simple: to be a space for social scientists like me (and maybe even others) to explore all of those Interesting Things in society, culture, public affairs, politics and be able to say scandalous phrases like ‘somewhat unlikely’ and ‘it depends’ about what we find. And still get to publish it for the world to see.
Unrelatedly, ‘somewhat unlikely’ is an answer option for many Likert scale questions on surveys. I wish there was a deeper connection here to the central premise of this newsletter, but it’s mostly a cool coincidence and not entirely unrelated since I’ll almost certainly write about survey data.
Who am I to trust your attention with?
That is an extremely fair question.
I’m a social scientist who believes in the power of data to persuade, inform, clarify — but also misinform, exaggerate, obfuscate, and most of all annoy. As my grad school mentor, Gary King, used to put it: data science “is not about the data.” It’s really about everything that happens before and after data is created. The data itself is just an annoying barrier between what you want to know about the and knowledge of that thing in your brain. This is the “power” of data that I believe in, and it’s an annoying, malevolent power.
Most of what I’ve learned about science and that I’ll be drawing on while writing this newsletter I learned while getting my Ph.D. in political science from Harvard University, during which I got to write on (what I think are) some interesting topics on media, politics, and public opinion in American democracy.
One fun fact from grad school: I once helped put together one of the world’s first studies of public opinion towards a, uh, particularly bad pandemic in recent history. Over the course of 48 hours, I led a team of some fifteen behavioral scientists around the world to figure out what a government agency in Italy should tell people who were, as you may recall, understandably freaking out about a rapidly spreading, deadly new virus. We ran the study, wrote it up — and it literally shaped a government’s communication policy during the worst public health emergency of our times. But since then, not a single academic journal wanted to publish this study, even though it got more citations in a few months than anything else I’ve ever done.
Another fun story: early in grad school, I ran a study with two brilliant social scientists — Chris Lucas and Kevin Munger — trying to test how “deepfakes”, a novel technology in 2020, might affect political attitudes in the context of an election. A the time of writing this post, the year of our lord 2023, the journal that accepted our paper, that notably uses technology from 2020, told us that it will likely take another 9 months to “actually” publish the paper. You heard this right: another 9 months to basically put up a PDF on a web site.1
These two stories drive home the main reason I’m joining the virtuous club of Dudes Starting Social Science Substacks2: to write timely, interesting, serious, and hopefully useful things about social science that’s not mind-numbingly hard-to-read, stupidly out-dated from years of peer review, or stuck behind an academic journal’s paywall.
At the time that I’m starting up this newsletter, during the day I help one of the oldest social research organizations in the United States do better research for government agencies, media organizations, and non-profits. I’ve also written a couple of EPs of a genre of music that my bandmates and I are still unable to define, if you’re into that kind of thing.
If this or anything else I’ve said sounds interesting to you, give my newsletter a regular read by hitting this little button here.
Yours Definitely,
Soubhik3
You should still read the paper because it’s cool. But also read: my co-author Kevin Munger’s paper on why this system of peer review is awful and why we should fix this. Or fellow social scientist Adam Mastroianni’s analysis on why we should all probably just do something else.
Come on, I’m not going to actually list them out. (They’re all actually pretty great).
You’re probably wondering how to say my name. Someone once thought my name (rapper / stripper name?) was “show bizz”, and ever since that’s how I help the prounciation-ally challenged remember my name (just replace the “z” with a “k”).