I didn’t much like it. This thing reeks of data dredging in every step; I don’t see why they controlled for birth place where you’d think that current residence would be much more relevant (Baltimore has rich and poor areas like most big cities; and if nothing else, it’d give you an idea of infection vectors if carriers cluster around the harbor or specific places); I find it odd that their WAIS subtest shows zero (0.0) decrease in the infected group while their weirdo IQ test I’ve never heard of shows a fall; and I’m not sure how convincing I find their mice models* - to what extent does it really mimick human infections with no apparent symptoms? It wouldn’t surprise me if, every time you gave mice a big injection of infectious organisms, their scores fell simply because you made them sick with something, so I’m not sure whether the mice experiment part is testing the right causal hypothesis (it might be testing ‘raging infections decrease cognitive performance’, not ‘this algae, and not other infectious agents, decreases cognitive performance’).
I would not be surprised if this never replicates.
* kudos to them for trying to experimentally test it, though
Good points all—I was hoping you’d show up. It’s odd enough though that I would be quite interested in any attempts at replication. Course, that might be coming from my interest in the evolutionary history and ecological role of a virus that can apparently infect organisms as different as blue-green algae and mammals.
Fulltext: https://pdf.yt/d/dr3uP9XOtT1BPimU / https://dl.dropboxusercontent.com/u/5317066/2014-yolken.pdf / http://libgen.org/scimag7/10.1073/pnas.1418895111.pdf
I didn’t much like it. This thing reeks of data dredging in every step; I don’t see why they controlled for birth place where you’d think that current residence would be much more relevant (Baltimore has rich and poor areas like most big cities; and if nothing else, it’d give you an idea of infection vectors if carriers cluster around the harbor or specific places); I find it odd that their WAIS subtest shows zero (0.0) decrease in the infected group while their weirdo IQ test I’ve never heard of shows a fall; and I’m not sure how convincing I find their mice models* - to what extent does it really mimick human infections with no apparent symptoms? It wouldn’t surprise me if, every time you gave mice a big injection of infectious organisms, their scores fell simply because you made them sick with something, so I’m not sure whether the mice experiment part is testing the right causal hypothesis (it might be testing ‘raging infections decrease cognitive performance’, not ‘this algae, and not other infectious agents, decreases cognitive performance’).
I would not be surprised if this never replicates.
* kudos to them for trying to experimentally test it, though
Good points all—I was hoping you’d show up. It’s odd enough though that I would be quite interested in any attempts at replication. Course, that might be coming from my interest in the evolutionary history and ecological role of a virus that can apparently infect organisms as different as blue-green algae and mammals.