Efficient markets and the replication crisis

It’s interesting how ‘efficiency’ also applies to non-financial things. This can be the tendency of the lines at food stores to have the same waiting time despite differences of line length (e.g. a long line of people with small orders who are paying in cash, versus someone with a huge order that is holding everyone up), so people instinctively arrange themselves in such a way that that the time spent in line is the same, without any coordination.

Thus, for the n+1 patron, it makes no difference which queue or line is chosen–either the short or long line. In the context of chaos theory, this can be thought of as a self-organizing system or spontaneous order.

Likewise, for a totally different example of efficiency, it came to light that ‘fitness influencer’ Mike Israetel’s PhD thesis, “The Interrelationships of Fitness Characteristics in Division 1 Athletes,” is total rubbish and of no scholarly value, after it was finally critiqued by Solomon Nelson some 12 years after its publication in 2013.

This sounds trivial in the grand scheme of things, but I think the implications are deeper. The video generated significant discussion on Reddit and follow-up videos dissecting the thesis and also the implications for academia in general, inviting the obvious question of why anyone didn’t bother to proofread the thesis. Or how does such slop make it through the cracks, or goes undetected for so long?

But the fact it hadn’t been read was an implicit admission by the ‘marketplace of ideas’ that it would likely be worthless, which indeed was confirmed to be the case. A doctorate thesis, even a worthless one, is a lot of work to read, and no one is going to put forth the effort unless they have some reasonable suspicion there is at least some value to be gleaned. The vast majority of these papers, especially in anything to do with ‘sports or athletics science,’ are justifiably unread, as a priori the scholarly value is presumed to be zero.

This is relevant to the so-called replication crisis. The replication crisis is also a feedback problem or self-reinforcing: the expectation is these papers are worthless and thus unread, and this inattention means bad papers are ignored and pile up, further souring people’s expectations, leading to more unread and bad papers.

There is also a common misconception that the awarding of a doctorate always signifies quality work or is an endorsement of its quality, when in reality, the quality depends heavily on the program, the subject matter, the school, the supervisor and other factors. This actually makes it less rigorous than commonly assumed, whereas at least passing the qualifiers assumes some objective yardstick of knowledge that one is assessed on.

Reading people’s experiences on Reddit in the context of the Nelson video, ‘shoddy dissertations’ are actually pretty common, which isn’t that big of problem given that most are unread anyway, and the awardee has no plans to pursue a research career. But it’s when the doctorate is used to burnish one’s intellectual reputation does it invite more scrutiny, and can backfire if the research isn’t up to par.

An example of this a while back on Twitter was when a certain popular political commentator displayed a link on his bio to his math thesis, which although perhaps outwardly impressive to someone unversed at math, upon closer inspection amounted to proving the binomial theorem, among other ‘groundbreaking’ results, inviting ridicule, which became more widespread even among laypeople after being assured by experts that the thesis was bad. To save face, he subsequently removed the link.