Yesterday Ralf brought up an excellent point on my post about blogging verse traditional publishing:
You ignore the thousands of printed copies of those journals in your stats.
Having not gone into the post blind, having done some research into publishing over the years, I knew that this is indeed correct BUT not too relevant. I promised Ralf some numbers to back up this assertion and here they are from a 2011 survey of where academics in the UK obtain journal articles and % of them that are digital :
Personal subscription: 4.7% (23.2% digital)
Library: 65.2% (93.5% digital)
Department/school: 4.5% (85.2% digital)
Repository (2.2% all digital)
Free web journal (9.2% all digital)
Copy from friend 5.6% (75.8% digital)
Interlibrary loan 2.1% (48% digital)
Authors website 1.5% (all digital)
Other website 2.7% (all digital)
Other 2.4% (85.7%)
Two key components to take away from this:
1. 85.7% of all journal articles are accessed in digital form (though some may be read in hardcopy after being printed out)
2. Journals that are accessed through personal subscriptions (society journals) are mainly obtained in hard copy form.
This means that for some society journals there are under reporting in access numbers BUT that is only based on paid subscribers to their hard copy journal. Most societies do not have a large number of subscribers AND there is no guarantee they actually read all the articles. In terms of views it is probably safe to say that online viewing metrics are accurate measurements of total views give or take +5-15%, each journal would need to be examined on its own. For example, the AAA’s American Anthropologists get 1.7 million online views a year and if all 10,000 members of the AAA read all articles (about 40 a year) it would only add about 20% to the total. A nice increase but not enough to significantly alter the numbers I present of blogs verse traditional publishing. This research is backed up by other studies (Journal Article Growth and Reading Patterns, it’s OA), (Electronic Journals and Changes in Scholarly Article Seeking and Reading, also OA) that show that this trend of digital access and access through libraries is growing.
That all being said, these studies also found that Humanities and Social Sciences tend to lag in digital use behind other research areas. So it could be that my numbers are off by a greater amount? However, the real question is, are hard copy views enough to dramatically alter the results I gave in the last post? It depends, in the case of AAA they have a much higher potential reach outside of their digital material. In the case of World Archaeology, not so much. Even doubling those numbers only ups their most read article over two years to 2400 view. John Hawks reached 9 times that everyday with his old RSS feed.
Yes, there is some fuzzyness in using digital views but for the most part they give a rough idea of views (views being different from impact which I will discuss in another post). Moreover, there has been a massive increase in the use of digital means as the primary way to obtain journal articles. In a couple of years it may be the only way.