I’m sure that you’re right that more effort is being put into teaching how to read literature, I’ve seen evidence of that myself. I know that many more universities offer library courses these days and of course, it’s only fair to say that many labs have formal and informal journal clubs to help early career researchers learn how to read a paper efficiently and critically.

For what it is worth, I am also seeing more and more professors explicitly teach “how to read a scholarly article” (or set of them!) – in the sciences and social sciences – and as a librarian I often teach an alternative reading approach as well – at least to undergraduate students, i.e., how to read for citations and what function they serve in the article.

What I mean to say is that I don’t think the fact that scientists don’t read linearly is a bad thing. Scientists often follow ideas through the literature rather than reading papers one at a time. I think that this kind of orthogonal reading is a product of the way that scientists are taught to think and a way to cope with the sheer volume of articles.

In terms of interpreting the studies. I do wonder if there might be a little bit of experimenter effect in some of the studies. After all, the idea that print is somehow better than digital appeals to many in the information industry. As I mentioned in my post. All the studies are about quality of retention when reading a simple, linear text and they don’t take into account the rapid way that many scientists screen for relevant information. That’s not meant to be a methodological criticism, but a concern about over-interpreting such results.