Thinking about scientific reproducibility helps us to think about the coronavirus.
The eminent reproducibility expert John Ioannidis, who in 2005 crystallized scientific awareness of the reproducibility crisis, has just written “A fiasco in the making? As the coronavirus pandemic takes hold, we are making decisions without reliable data”—an article that queries whether we are making vast, economy- and society-wrecking decisions on what amounts to a false positive—insufficient data, badly analyzed, about infection rates and mortality. Ioannidis is generalizing his experience of the last decades of biomedical research, where the argument has always been:
This is an emergency, we need to act or people will die, we must immediately use this partially tested serum—oops, guess we wasted $10 billion on a placebo.
Ioannidis is afraid that we are about to jump off a cliff on the basis of insufficient data, and with no clear idea of the landscape below.
This is all true. And the news today is that there are 3,405 dead in Italy, and counting. The arguments in favor of acting now, with insufficient data, are unusually strong.
Right now, the most important thing we can do as citizens is to keep questions of reproducibility in mind—to be a little skeptical of the latest headline, to ask can this data be reproduced, is the statistical analysis appropriate, are the journalists chasing after an exciting false positive? Instant credulity of the latest headlines is no more appropriate now than at any other time.
On the other hand, as the old saw says, If you can keep your head while all about you are losing theirs, you probably don’t know what’s going on. The Greeks tell us to prefer moderation to either extreme: we should seek the middle road between panic and stubborn inertia.
In the longer run, Americans, rightly, are already beginning to think about what policy is appropriate for the medium and long term—How do we contain recurrence of the coronavirus? How do we prepare for the next pandemic? Unfortunately, we are likely to hard-wire policy now, based on our insufficient data and possibly faulty analyses. Ioannidis’ caution, which asks for transparent science and reproducibility before we act, is entirely appropriate as we consider what spending decisions, laws, and regulations to undertake. Every lawmaker, legislative assistant, administrator, and federally employed scientific expert should read Ioannidis’ article, and keep his cautions in mind.
And every professor, every teacher, should be on the alert against the descent of a new political correctness and groupthink about the coronavirus, about epidemiology, about appropriate public policy. We should expect some new intolerance, aimed at coronavirus denialists, probably allied with a fanatical insistence to use politically correct language about the Wuhan coronavirus. Higher education’s default, alas, is now for intolerant groupthink. If allowed to extend to discussion of the science of the coronavirus, it may have immediately deadly consequences.
We must be prepared now to champion open discussion about the science of the coronavirus, and about all related public policy.