Post navigation

MSc and PhD programs in statistics at the University of British Columbia

A young economist who prefers to remain anonymous writes:

Francesca Gino, the Harvard behavioural professor accused of fraud has had her tenure revoked from Harvard and her employment terminated. I’m honestly a bit surprised that this has happened, given how we’ve seen similar researchers facing similar accusations receive not even a slap on the wrist! But given the overwhelming evidence against her, this has to be a positive development, right? Hopefully one that leads to more similar such punishments?

Side note, would you be interested in hearing about another paper that claims to measure one thing but probably doesn’t? There’s a an interesting paper trying to quantify the impact of religion /religious belief on 17th-19th century development of science that seems to me to be using many proxy variables that are either not measuring what the author claims or have potentially such biased measurement errors that I don’t trust them much.

Regarding the first item, I replied that it says it was the “first time Harvard has removed a tenured instructor in about 80 years.” But they got rid of psychology professor Marc Hauser in 2012 for misreporting his monkey data. Maybe that’s different because they didn’t fire him, they just forced him to quit? My correspondent replied:

Yes, my limited understanding is that Hauser quit before anything else was done. Possibly forced to resign? I don’t know the details. Whereas Gino did not resign, she lost her tenure and so could be fired. An article by the Harvard Crimson suggests this may be unprecedented, all previous professors in similar situations having resigned.

I guess that Gino was different because most of the people involved in these scandals seem to want the scandal to go away quietly, and resigning is part of that. If the university wants to get rid of you, then your resignation is a bargaining chip that you can use to get better treatment, and the endgame is that you leave with a minimum of fuss. Gino is different because with her lawsuit she expressed a comfort with not going away quietly. If she wasn’t going to leave on her own, or if her conditions for retirement were to ask for $50 million or whatever, then it makes sense that Harvard just get rid of her. I say all this speculatively, not knowing any of the people involved. (I did have a brief exchange with Gino once, but all of it concerned statistics and research methods; Harvard and the lawsuit did not come up.)

Regarding the second item, my correspondent elaborates:

The paper is pretty ambitious and I appreciate what it’s trying to do. Trying to systematically quantify scientific accomplishments and the religiosity of scientists is an incredibly difficult task. But I’m really not sure that the paper succeeded.

Just as one (important) example, in section 2.3 the paper uses different measures of “impact of scientist” for individual level regressions. What are these measures? They’re taken from a different paper (Laouenan et al., 2022) and are all basically Wikidata/Wikipedia measures: Word count of all biographies, Biography readers and Number of Wikipedia editions.

There’s so many issues with doing this, I would not treat any of these measures as good examples of the “impact” (already a somewhat nebulous term) a scientist had. What exactly are we capturing by looking at the amount of focus devoted by 21st century Wikipedia writers and readers towards any individual scientist by these 3 measures? I’m really not sure, but I’m dubious that it really correlates well with the idea of the importance of a scientist had for the development of science. Furthermore, these measures will likely have measurement errors…

And I’d presume these measurement errors are likely correlated with the scientists religiosity (which also have measurement errors of their own); the paper says as much “First, could the excessive fame of these freethinkers be related not to their intellectual contributions but instead to the controversies surrounding their religious views?” which is only of many channels through which there could be endogeneity here.

Just as another example, since most Wikipedia writers are from rich countries, with much more atheistic populations compared to the rest of the world, it’s not a stretch to assume they are more likely to devote attention towards atheistic scientists from previous centuries. Ditto for readers.

They try to mitigate things as best they can, but overall I’m rather sceptical of the results. Like I said, an ambitious paper, but one that, at least to me, seems to not do what it claims it does.

The paper, Science and Religious Dogmatism, is by Matías Cabello, and its abstract concludes:

Throughout modern Western history, and within a given city and time period, scientists who doubted God and the scriptures have been considerably more productive than those with dogmatic beliefs.

I agree with my anonymous correspondent that such things are hard to measure, not just scientific productivity or importance or influence but also religious beliefs. I’ve often wondered what my religious beliefs would’ve been, had I grown up among my ancestors 200 or 300 or 400 years ago. (I have a pretty good idea what my beliefs would’ve been had I been born into my family 100 years ago, because that was around when my parents were growing up.)

P.S. It’s interesting that the author of the paper is studying economics, as this sort of thing seems like straight-up sociology. I guess there are just lots more economists than sociologists out there doing historical and quantitative studies.