How does PLOS ONE maintain its impact factor?
There are, I think, two distinct factors at work that may help explain some of your puzzlement:
- Your field's impact factor is not academia's impact factor. For example, society journals in my field have an impact factor of ~ 5, and some of the big names for very splashy studies have impact factors ranging from 20 to 56. Depending on the balance of fields submitting to PLOS ONE, their impact factor may be coming from more cited fields.
- Long-tailed citation papers. Impact factors, like many averages, are susceptible to long-tail effects. PLOS ONE is an open-access journal, and a highly visible one. It's possible that the occasional highly accessible generalist paper makes it there, and yields a large number of citations as a result, pulling up the overall impact factor.
Both of these are helped, in my opinion, by the lack of review for "importance" - beyond your suggestion that this does result in less important papers ending up in PLOS ONE, it's also a benefit to papers that don't quite "fit" in highly specialized journals, but may still be impactful.
I think that PLOS ONE is gambling on two key hypotheses:
- People are very bad at judging future importance --- thus, no "significance" filtering.
- Search engines and social networks are now much better at delivering articles than subscriptions --- thus, open access.
This certainly conforms with my experience: at present I have two PLOS ONE articles, each published about 5 years ago, one with 80 citations and the other with 9 citations as of this writing. Both are quite specialized and likely would have had a hard time getting published in "selective" journals, yet have found some sort of audience. I'm thus not surprised that they seem to be able to maintain a reasonable impact factor.
Note this answer is anecdotal, but since I have heard this stated by 3 high volume publishing professors now, I believe that it may be a factor. These 3 professors claimed to rank journals as follows (approximately)
- nature and science
- PNAS and the top specialized (but still fairly broad) Journal in their field
- All other journals including PLOS one
One philosophy for academic success is that once you have many publications in reasonably good journals, the marginal benefit of one more isn't that great. However, the marginal benefit of one more Science or Nature paper is quite big. Therefore, since PLOS one is often less of a hassle to submit to, if a paper is rejected from Science or Nature or PNAS the next stop for these high impact professors is often PLOS one. It just isn't worth it for these busy professors to trudge through multiple submissions down the journal food chain because they want to focus on their next Science/Nature submission or Grant proposal which could lead to such a paper. This means PLOS one gets a lot of papers that are written by pretty famous professors that just weren't jazzy enough for Nature and Science. Of course, PLOS one also gets a bunch of unimportant papers (by famous and non-famous authors) but I suspect some of the heavy tail papers that @fomite suggested are from academics with this philosophy.
I again stress that this is anecdotal and that I have no data to confirm this hypothesis.