Anthony Point 1 - absolutely true. Only a small minority of downloads lead to citations. Have a look at the download data of eprints.utas.edu.au. However I cannot resist writing that citations are not the same as impact. Only in academic circles are citations highly regarded and that is useful but not totally relevant. In most places what one is looking for is impact. For example if I write an article on improving fish farming, I want the fish farmers to take it up. If they only get the AM, well yes that is much better than nothing (as Stevan says) and they might contact me. However, their directors will want to know (for legal reasons) that their recommendations are firmly-based, and that means access to the VoR, which fish farmers may not have. Hence researcher attitudes to the VoR.
Point 2 - sorry no. The observation is anecdotal. Largely based on my university and Australian universities, but supported by website, blog and Mendeley evidence. I believe it is why some mandates are not worth their disk space (or the paper they may be written on) - they are ignored by real live researchers. The OA movement needs to engage with researchers and convince them that the mandate is worth complying with, because they do not believe it. You may be interested to know that ALL Australian universities have repositories, but only those of the University of Queensland, the Queensland University of Technology, and the University of Tasmania (mine) are in the top 100 of the Webometrics survey? Why? Probably because most of the 'mandates' are ineffective, except in gathering citations and restricted documents. The first two universities have strong mandates. The OA world is bigger than 'mandates at all costs'. It needs to recognise the reality of revolutions. They disrupt normal practice (in this case of science and scholarly dissemination). Best wishes Arthur -----Original Message----- From: goal-bounces at eprints.org [mailto:goal-boun...@eprints.org] On Behalf Of Andrew A. Adams Sent: Wednesday, 15 February 2012 6:47 PM To: Global Open Access List (Successor of AmSci) Subject: [GOAL] Re: Author's refereed, revised, accepted final draft vs. publisher's version-of-record In response to Stevan Harnad, Arthur Sale wrote: > When we turn to the researcher, the situation changes significantly, > if slightly. Researchers regard the VoR as the canonic version of > their article, almost exclusively (I exempt you and me and a small set > of similar-minded people). As far as they are concerned, all earlier > versions are suspect and not to be displayed once they have served > their purpose. They also believe they wnthe VoR. This is not an > cademic ideal but a practical reality. The VoR is THE CANONIC > VERSION. It is one reason why many researchers fail to post anything > on an OA repository, because they do not understand what their rights > are and they are reluctant to post something they conceive of as > flawed. There's an assumption in many of the posts on this topic that all articles accessed will be cited. My experience is that I identify many articles from their abstract (usually available for free), a forward and backwards reference search (an article is cited by another I've read or cites another one I've read), from the list of publications of an author whose other works I've read and from a number of other sources. If that article is available to me in the VoR or as an AM then I can first skim the introduction/conclusions and if it seems of further interest read the full article, or selected elements of it. After this proper reading of all or some of either the VoR or the AM then at some point I MAY wish to reference the article or quote from it. Then and only then is the VoR actually needed at all, ad actually I (as you note below) rely on the open access AM version if I don't have access already to the VoR (of course any article I don't have access to doesn't get read and therefore not cited - in particular I almost never pay the ridiculous per-article costs requested by publishers - one article costing the same as 50-100% of full books? That just demonstrates exactly how ridiculous are the subscription rates on which the per-article charges are sert pro-rata). If I really felt I needed the VoR for the articles I want to cite then I could pay the per article charge (I don't, but others may be more hesitant). In my experience, and this is just personal anecdote, I identify perhaps 50-100 times as many articles as of potential interest as I actually cite. For someone in a less interdisciplinary field perhaps their numbers might be lower, but then again they may also already have subscription access to the journals they feel they need - the narrower one's research focus, and the large one's group of researchers with the same interest, the more likely one is to have access to the necessary literature. However, I would suspect that most researchers do not cite every article they ever read. For any article one does not actually wish to cite, the VoR is not necessary. The AM should absolutely be sufficient for evaluating the importance of the article. Arthur Sale continued: > Interestingly though, I believe there are a growing number of > researchers who totally ignore any agreement they sign with > publishers, and post their VoR regardless, because it is heirs It is > this practice (in the form of providing electronic "reprints") that > publishers find difficult to ignore, and possibly why the copyright > transfer agreements are strengthened. [snip] You make a quantitative claim here. Do you have any evidence you can offer for this? [further snippage] -- Professor Andrew A Adams aaa at meiji.ac.jp Professor at Graduate School of Business Administration, and Deputy Director of the Centre for Business Information Ethics Meiji University, Tokyo, Japan http://www.a-cubed.info/ _______________________________________________ GOAL mailing list GOAL at eprints.org http://mailman.ecs.soton.ac.uk/mailman/listinfo/goal
