Open Access Archivangelism

What follows below is what we hope will be found to be a conscientious and attentive series of responses to questions raised by Phil Davis about our paper (Gargouri et al, currently under refereeing) -- responses for which we did further analyses of our data (not included in the draft under refereeing).

Gargouri, Y., Hajjem, C., Lariviere, V., Gingras, Y., Brody, T., Carr, L. and Harnad, S. (2010) Self-Selected or Mandated, Open Access Increases Citation Impact for Higher Quality Research.(Submitted)

We are happy to have performed these further analyses, and we are very much in favor of this sort of open discussion and feedback on pre-refereeing preprints of papers that have been submitted and are undergoing peer review. They can only improve the quality of the eventual published version of articles.

However, having carefully responded to Phil's welcome questions, below, we will, at the end of this posting, ask Phil to respond in kind to a question that we have repeatedly raised about his own paper (Davis et al 2008), published a year and a half ago...

RESPONSES TO DAVIS'S QUESTIONS ABOUT OUR PAPER:

PD:
"Stevan, Granted, you may be more interested in what the referees of the paper have to say than my comments; I'm interested in whether this paper is good science, whether the methodology is sound and whether you interpret your results properly."

We are very appreciative of your concern and hope you will agree that we have not been interested only in what the referees might have to say. (We also hope you will now in turn be equally responsive to a longstanding question we have raised about your own paper on this same topic.)

PD:
"For instance, it is not clear whether your Odds Ratios are interpreted correctly. Based on Figure 4, OA article are MORE LIKELY to receive zero citations than 1-5 citations (or conversely, LESS LIKELY to receive 1-5 citations than zero citations). You write: "For example, we can say for the first model that for a one unit increase in OA, the odds of receiving 1-5 citations (versus zero citations) increased by a factor of 0.957. Figure 4.. (p.9)"

You are interpreting the figure incorrectly. It is the higher citation count that is in each case more likely, as co-author Yassine Gargouri pointed out to you in a subsequent response, to which you replied:

PD:
"Yassine, Thank you for your response. I find your odds ratio methodology unnecessarily complex and unintuitive but now understand your explanation, thank you."

Our article supports its conclusions with several different, convergent analyses. The logistical analysis with the odds ratio is one of them, and its results are fully corroborated by the other, simpler analyses we also reported, as well as the supplementary analyses we append here now.

[Yassine has since added that your confusion was our fault because we had used the value 0.957 by way of an illustration. We should have chosen a better example, where (Exp(ß)) is clearly greater than 1; the value 0.957 is below 1, and too close to 1, to serve as an illustration. We should have said: "For the second model, a one-unit increase in OA, the odds of receiving 5-10 citations (versus zero 1-5 citations) increased by a factor of 1.323." This clearer example will be used in the revised text of the paper.]

PD:
"Similarly in Figure 4 (if I understand the axes correctly), CERN articles are more than twice as likely to be in the 20+ citation category than in the 1-5 citation category, a fact that may distort further interpretation of your data as it may be that institutional effects may explain your Mandated OA effect. See comments by Patrick Gaule and Ludo Waltman on the review"

Here is the analysis underlying Figure 4, re-done without CERN, and then again re-done without either CERN or Southampton. As will be seen, the outcome pattern, as well as its statistical significance, are the same whether or not we exclude these institutions. (Moreover, I remind you that those are multiple regression analyses in which the Beta values reflect the independent contributions of each of the variables: That means the significant OA advantage, whether or not we exclude CERN, is the contribution of OA independent of the contribution of each institution.)

SUPPLEMENTARY FIGURE S1

PD:
"Changing how you report your citation ratios, from the ratio of log citations to the log of citation ratios is a very substantial change to your paper and I am surprised that you point out this reporting error at this point."

As noted in Yassine's reply to Phil, that formula was incorrectly stated in our text, once; in all the actual computations, results, figures and tables, however, the correct formula was used.

PD:
"While it normalizes the distribution of the ratios, it is not without problems, such as: 1. Small citation differences have very large leverage in your calculations. Example, A=2 and B=1, log (A/B)=0.3"

The log of the citation ratio was used only in displaying the means (Figure 2), presented for visual inspection. The paired-sample t-tests of significance (Table 2) were based on the raw citation counts, not on log ratios, hence had no leverage in our calculations or their interpretations. (The paired-sample t-tests were also based only on 2004-2006, because for 2002-2003 not all the institutional mandates were yet in effect.)

Moreover, both the paired-sample t-test results (2004-2006) and the pattern of means (2002-2006) converged with the results of the (more complicated) logistical regression analyses and subdivisions into citation ranges.

PD:
"2. Similarly, any ratio with zero in the denominator must be thrown out of your dataset. The paper does not inform the reader on how much data was ignored in your ratio analysis and we have no information on the potential bias this may have on your results."

As noted, the log ratios were only used in presenting the means, not in the significance testing, nor in the logistic regressions.

However, we are happy to provide the additional information Phil requests, in order to help readers eyeball the means. Here are the means from Figure 2, recalculated by adding 1 to all citation counts. This restores all log ratios with zeroes in the numerator (sic); the probability of a zero in the denominator is vanishingly small, as it would require that all 10 same-issue control articles have no citations!

The pattern is again much the same. (And, as noted, the significance tests are based on the raw citation counts, which were not affected by the log transformations that exclude numerator citation counts of zero.)

SUPPLEMENTARY FIGURE S2
This exercise suggested a further heuristic analysis that we had not thought of doing in the paper, even though the results had clearly suggested that the OA advantage is not evenly distributed across the full range of article quality and citeability: The higher quality, more citeable articles gain more of the citation advantage from OA.

In the following supplementary figure (S3), for exploratory and illustrative purposes only, we re-calculate the means in the paper's Figure 2 separately for OA articles in the citation range 0-4 and for OA articles in the citation range 5+.

SUPPLEMENTARY FIGURE S3:
The overall OA advantage is clearly concentrated on articles in the higher citation range. There is even what looks like an OA DISadvantage for articles in the lower citation range. This may be mostly an artifact (from restricting the OA articles to 0-4 citations and not restricting the non-OA articles), although it may also be partly due to the fact that when unciteable articles are made OA, only one direction of outcome is possible, in the comparison with citation means for non-OA articles in the same journal and year: OA/non-OA citation ratios will always be unflattering for zero-citation OA articles. (This can be statistically controlled for, if we go on to investigate the distribution of the OA effect across citation brackets directly.)

PD:
"Have you attempted to analyze your citation data as continuous variables rather than ratios or categories?"

We will be doing this in our next study, which extends the time base to 2002-2008. Meanwhile, a preview is possible from plotting the mean number of OA and non-OA articles for each citation count. Note that zero citations is the biggest category for both OA and non-OA articles, and that the proportion of articles at each citation level decreases faster for non-OA articles than for OA articles; this is another way of visualizing the OA advantage. At citation counts of 30 or more, the difference is quite striking, although of course there are few articles with so many citations:

SUPPLEMENTARY FIGURE 4

---

REQUEST FOR RESPONSE TO QUESTION ABOUT DAVIS ET AL'S (2008) PAPER:

Davis, PN, Lewenstein, BV, Simon, DH, Booth, JG, & Connolly, MJL (2008)
Open access publishing, article downloads, and citations: randomised controlled trial British Medical Journal 337

Critique of Davis et al's paper: "Davis et al's 1-year Study of Self-Selection Bias: No Self-Archiving Control, No OA Effect, No Conclusion" BMJ Responses.

Davis et al had taken a 1-year sample of biological journal articles and randomly made a subset of them OA, to control for author self-selection. (This is comparable to our mandated control for author self-selection.) They reported that after a year, they found no significant OA Advantage for the randomized OA for citations (although they did find an OA Advantage for downloads) and concluded that this showed that the OA citation Advantage is just an artifact of author self-selection, now eliminated by the randomization.

What Davis et al failed to do, however, was to demonstrate that -- in the same sample and time-span -- author self-selection does generate the OA citation Advantage. Without doing that, all they have shown is that in their sample and time-span, they found no significant OA citation Advantage. This is no great surprise, because their sample was small and their time-span was short, whereas the many of the other studies that have reported finding an OA Advantage were based on much larger samples and much longer time spans.

The question raised was about controlling for self-selected OA. If one tests for the OA Advantage, whether self-selected or randomized, there is a great deal of variability, across articles and disciplines, especially for the first year or so after publication. In order to have a statistically reliable measure of OA effects, the sample has to be big enough, both in number of articles and in the time allowed for any citation advantage to build up to become detectable and statistically reliable.

Davis et al need to do with their randomization methodology what we have done with our mandating methodology, namely, to demonstrate the presence of a self-selected OA Advantage in the same journals and years. Then they can compare that with randomized OA in those same journals and years, and if there is a significant OA Advantage for self-selected OA and no OA Advantage for randomized OA then they will have evidence that -- contrary to our findings -- some or all of the OA Advantage is indeed just a side-effect of self-selection. Otherwise, all they have shown is that with their journals, sample size and time-span, there is no detectable OA Advantage at all.

What Davis et al replied in their BMJ Authors' Response was instead this:

PD:
"Professor Harnad comments that we should have implemented a self-selection control in our study. Although this is an excellent idea, it was not possible for us to do so because, at the time of our randomization, the publisher did not permit author-sponsored open access publishing in our experimental journals. Nonetheless, self-archiving, the type of open access Prof. Harnad often refers to, is accounted for in our regression model (see Tables 2 and 3)... Table 2 Linear regression output reporting independent variable effects on PDF downloads for six months after publication Self-archived: 6% of variance p = .361 (i.e., not statistically significant)... Table 3 Negative binomial regression output reporting independent variable effects on citations to articles aged 9 to 12 months Self-archived: Incidence Rate 0.9 p = .716 (i.e., not statistically significant)..."

This is not an adequate response. If a control condition was needed in order to make an outcome meaningful, it is not sufficient to reply that "the publisher and sample allowed us to do the experimental condition but not the control condition."

Nor is it an adequate response to reiterate that there was no significant self-selected self-archiving effect in the sample (as the regression analysis showed). That is in fact bad news for the hypothesis being tested.

Nor is it an adequate response to say, as Phil did in a later posting, that even after another half year or more had gone by, there was still no significant OA Advantage. (That is just the sound of one hand clapping again, this time louder.)

The only way to draw meaningful conclusions from Davis et al's methodology is to demonstrate the self-selected self-archiving citation advantage, for the same journals and time-span, and then to show that randomization wipes it out.

Until then, our own results, which do demonstrate the self-selected self-archiving citation advantage for the same journals and time-span, show that mandating the self-archiving does not wipe it out.

Meanwhile, Davis et al's finding that although their randomized OA did not generate a citation increase, it did generate a download increase, suggests that with a larger sample and time-span there may well be scope for a citation advantage as well: Our own prior work and that of others has shown that higher early download counts tend to lead to higher later citation counts.

Bollen, J., Van de Sompel, H., Hagberg, A. and Chute, R. (2009) A principal component analysis of 39 scientific impact measures in PLoS ONE 4(6): e6022,

Brody, T., Harnad, S. and Carr, L. (2006) Earlier Web Usage Statistics as Predictors of Later Citation Impact. Journal of the American Association for Information Science and Technology (JASIST) 57(8) 1060-1072.

Lokker, C., McKibbon, K. A., McKinlay, R.J., Wilczynski, N. L. and Haynes, R. B. (2008) Prediction of citation counts for clinical articles at two years using data available within three weeks of publication: retrospective cohort study BMJ, 2008;336:655-657

Moed, H. F. (2005) Statistical Relationships Between Downloads and Citations at the Level of Individual Documents Within a Single Journal. Journal of the American Society for Information Science and Technology 56(10): 1088- 1097

O'Leary, D. E. (2008) The relationship between citations and number of downloads Decision Support Systems 45(4): 972-980

Watson, A. B. (2009) Comparing citations and downloads for individual articles Journal of Vision 9(4): 1-4

University of Strathclyde
and
Brunel University

---

Please register your own university's mandate in ROARMAP too, to track progress and to encourage other universities to adopt mandates of their own.

---

SUMMARY: The Association of Universities in the Netherlands (VSNU) has made a deal with Springer that articles by VSNU authors will be made OA. But Springer is already on the side of the angels on OA, being completely Green on immediate, unembargoed author OA self-archiving. Hence all VSNU authors are already free to deposit their refereed final drafts of their Springer articles in their institutional repositories, without requiring any further permission or payment. So what in addition is meant by the VSNU deal with Springer? that the Springer PDF rather than the author's final draft can be deposited? That Springer does the deposit on VSNU authors' behalf? Or is this a deal for prepaid hybrid Gold OA? In the case of Springer articles, it seems that what the Netherlands lacked was not the right to make them OA, but the mandate (from the VSNU universities and Netherlands' research funders like NWO) to make them OA. There are some signs, however, that this too might be on the way...

---

In a press release entitled "Dutch higher education sector convinced of need for Open Access," the SURF Foundation in the Netherlands wrote:

"The Association of Universities in the Netherlands (VSNU) has reached agreement with Springer that in 2010 all articles by Dutch researchers in Springer journals will be made available Open Access, subject to the author agreeing. Other publishers too are providing opportunities for Open Access publication because they are following Springer in allowing researchers to arrange for Open Access when publishing their articles. Almost all publishers already allow researchers to upload the definitive author's version of their article to their institution's repository."

It would be very helpful if SURF or VSNU could explain a little more clearly what this means:

(1) Is it that VSNU has made a deal with Springer (as University of California has done) that articles by VSNU authors will be made OA?

(2) How will those articles be made OA?

Springer is already on the side of the angels, being completely Green on immediate, unembargoed author OA self-archiving. In other words, VSNU authors are all already free to deposit their refereed final drafts of their Springer articles in their institutional repositories, without requiring any further permission or payment.

Hence it is unclear what, over and above this, is meant by (1)? that the Springer PDF rather than the author's final draft can be deposited? That Springer does the deposit on author's behalf? Or is this a deal for prepaid hybrid Gold OA?

It is important to raise these questions, because in the case of Springer articles, it seems that what the Netherlands lacked was not the right to make them OA, but the mandate (from the VSNU universities and Netherlands' research funders like NWO) to make them OA.

"One problem for scientists and scholars is the need to publish in prestigious and expensive journals so as to receive a good rating, which is important when applying for grants from organisations such as the NWO. Prof. Engelen said that the NWO would investigate ways of ensuring that publications in Open Access would count more significantly towards the author's 'impact factor.'"

Does this mean that Springer articles should now count more for NWO than they do now? Why? Should it not be the quality standards of each journal that determine how much it counts for NWO? (And also, of course, the citation impact of each article itself.)

Is being OA supposed to make an article count more? Why? (Especially since making an article OA has already been shown to increase its citation impact?)

Is this not the usual error, of assuming that "OA" means "published in a Gold OA journal" -- and assuming also that Gold OA journals are new journals, and have to compete with established journals in order to demonstrate their quality standards?

If so, why should any journal count more just because it is Gold OA?

And what about Green OA, which any Netherlands author can already provide for their articles, and especially with Springer articles, which already have Springer's endorsement for Green OA?

Green OA is already based on each journal's quality standards and track-record. No special preferential treatment is required.

"Paul Doop – a member of the board of Amsterdam University and Amsterdam University of Applied Sciences, and chair of the ICT and Research platform board of SURFfoundation – argued that the problem could be solved by including a provision for mandatory Open Access in collective labour agreements."

This is certainly one possible way to mandate OA. Or, better, each VSNU university could simply adopt a policy, as over 100 universities worldwide have already done, that requires the deposit of all institutional refereed research output in the institution's repository.

But this has nothing whatsoever to do with the "problem" of making new Gold OA journals "count" more than they have earned with their quality standards, just as every other journal has done. Indeed, mandating Green OA has nothing to do with Gold OA journals at all (except that all Gold OA journals are also Green!)

"Many of those attending the seminar thought that was going too far. Prof. Engelen said, however, that his organisation was keeping close track of developments and that if insufficient progress had been made in a year’s time, the NWO would see whether it could make Open Access obligatory, as its sister organisations in the United Kingdom and the United States have already done."

This would be splendid. And I hope NWO will not wait so long to do what the US and UK (and many other countries) are already doing.

But it would be helpful if the very timely and commendable plan to mandate Green OA in the Netherlands is not conflated with the completely different question of paying for Gold OA, or with trying to make Gold OA journal articles "count" more.

Stevan Harnad
American Scientist Open Access Forum

SANDY THATCHER: "it's the peer review that is the most expensive part of the whole process, and arXiv is not in the business of peer reviewing."

DAVID PROSSER:: "Is that true, Sandy? Can we have a reference please? Tenopir and King back in 2004 suggested that 'manuscript receipt processing, disposition decision-making, identifying reviewers or referees and review processing' constituted 26% of the direct costs of producing an article (which they estimated at $1700 on average). Of course, costs may have shifted in the years since then. Which is why a reference would be welcome."

What Sandy Thatcher said is perfectly correct:

(1) The cost of providing peer review (c. $500 per article -- though more efficient online procedures could lower that) is indeed the most expensive part of the process of providing a peer-reviewed article for free (OA) by depositing it in a central repository like Arxiv (or in the author's own Institutional Repository, IR).

(2) And Arxiv does not provide the peer review. (Nor does any other repository.)

(3) Low as it is, $7 per article just for deposit and archiving is probably an overestimate, because Arxiv needs to do far too much work to process and store all the world's institutions' physics deposits centrally: It would cost even less per article for an Institutional Repository (IR) that archives only its own annual research output (and knows all its own researchers, hence need not do the extra generic precautionary controls). (Be careful not to jig the estimate by factoring in the costs of online infrastructure that the institution already has, regardless of whether it has an IR: just the one-time IR set-up cost, the extra server and disk-space, etc., plus the cost per deposit and annual maintenance of the IR only.)

It would be useful to have IRs' estimates of their annual cost per article deposited -- but only from mature mandated IRs that are already well on the way to capturing 100% of their annual institutional output of refereed journal articles. (Obviously the IR price per article will be somewhat higher for IRs that are still only capturing only 15% or less of their annual refereed research output, as most IRs today still are, because they have not yet mandated deposit.)

Another useful comparison would be the cost -- in money and time -- of doing the unnecessary IR "quality controls" and preprocessing that many IRs think, superstitiously and superfluously, that they need to do. (In this case, estimates from all the immature, near-empty IRs are relevant too.)

At Southampton ECS, the first mandated IR of all (since 2002), we realized within the first year of the mandate that the "quality control" (for the content and metadata of the deposit) was based on a completely unnecessary and dysfunctional misanalogy with library collections and cataloguing, that all it did was create needless work and backlogs for the "quality-controllers" and needless resistance and counterproductive resentment from depositing authors who, having taken the trouble to deposit their refereed final drafts, as mandated, were then denied the immediate satisfaction of seeing their deposits go immediately online and start getting downloaded: instead, they had to go into a quality-control queue, sometimes for days or weeks, as the volume of mandated deposits to "process" grew. We quickly jettisoned the gratuitous process and have seen the IR's deposits growing happily ever since.

Leave any "quality control" for your institutional authors' peer-reviewed final drafts in the background. If something is wrong, users will let the author know; if users don't squawk (or there are no users!), the slip-up probably isn't even worth correcting. Focus on solving the real problem, which is not "quality control" but capturing the IR's target content: the institution's full annual output of refereed research.

And remember that -- whilst journals still exist and subscriptions are still paying for their quality control -- your IR is not hosting the all-important version-of-record, but merely an OA supplement.

A word to the wise...

Stevan Harnad
American Scientist Open Access Forum

Colin G. Scanes Editor-in-Chief Poultry Science (Poultry Science Association) wrote:

[T]here is self-interest from journals... whether... profit... or non-profit... in not supporting free open access. Equally there is strong self-interest in university libraries... supporting open access because they are likely to reduce their costs of purchasing journals.

-- There are also the interests of research, researchers, their institutions, their funders, and the tax-paying public that supports the research and for whose benefit it is conducted and published. That interest is in making the research accessible, immediately upon acceptance for publication, to all would-be users, not just those whose institutions can afford subscription access.

Hitchcock, S. (2010) The effect of open access and downloads ('hits') on citation impact: a bibliography of studies

1. Who is to pay the very real costs of producing journals with this move to open access? Should it be the researcher, and, if so, where is the additional funding to come from? Is it realistic to consider that journals should absorb the costs

-- Open Access means free online access to published journal articles, not necessarily Open Access publishing. Authors can provide Open Access to their conventionally published articles by self-archiving their final refereed drafts free for all online.

2. At what point do libraries cease to purchase subscriptions for journals if their contents are available by open access?

-- No one knows whether and when libraries will cancel journals. Till they do, institutional subscriptions pay the cost of peer review and authors make their final drafts free for all online. If and when journal cancellations make subscriptions unsustainable because users prefer to use the free online drafts, journals will cut costs and downsize to providing peer review alone, paid for, per article, by authors' institutions, out of their windfall subscription cancellation savings.

Harnad, S. (2007) The Green Road to Open Access: A Leveraged Transition. In: The Culture of Periodicals from the Perspective of the Electronic Age, pp. 99-105, L'Harmattan.

3. If library subscriptions to journals are an essential part of the business plan of a journal or a professional society, how many journals will disappear if we go to a completely open access approach?

-- No journals will disappear as a result of Open Access. Open Access is provided by author self-archiving (now being increasingly mandated by their institutions and funders) and if and when subscriptions fail, journals will downsize to peer-review service provision alone, paid for on the open access publishing service-fee model.

4. As a journal editor with, at present, a positive cash flow, we can and do waive page charges from papers from institutions in developing countries that cannot afford to pay these. We will not be able to continue this if there is a major reduction in revenue. Forcing journals to adopt an author-pays model would have a stifling effect on the publication of work from authors in developing countries.

-- No need to change anything (except to make sure the journal endorses rather than obstructs author self-archiving). Universal self-archiving and self-archiving mandates will provide universal Open Access, and the rest depends on how long subscriptions remain sustainable, and on whether and when the downsizing and transition to the Open Access cost-recovery model occurs.

5. What is a reasonable embargo period between publication and the paper being available by free open access?

-- What is optimal for research -- and for researchers, their institutions, their funders, and the tax-paying public that supports the research and for whose benefit it is conducted and published -- is no embargo at all. What is helpful from publishers is if they endorse Open Access self-archiving by authors. The rest will all come as a natural matter of course either way (i.e., with or without publisher endorsement), as a result of Open Access mandates by institutions and funders. The Green publishers will simply have the historic satisfaction of having been on the side of the angels all along.

Poultry Science's self-archiving policy is not in Romeo and does not appear to be among the 63% of journals that endorse immediate Open Access self-archiving by its authors. It would be helpful if this were remedied:

Poultry Science Copyright Release: Copyright laws make it necessary for the Association to obtain a release from authors for all materials published. To this end we ask you to grant us all rights, including subsidiary rights, for your article. You will hereby be relinquishing to the Poultry Science Association all control over this material such as rights to make or authorize reprints, to reproduce the material in other Association publications, and to grant the material to others without charge in any book of which you are the author or editor after it has appeared in the journal.

Stevan Harnad
American Scientist Open Access Forum

Nat Gustafson-Sundell wrote:

NGS: "I don't expect local repositories to ever offer quality control."

Of course not. They are merely offering a locus for authors to provide free access to their preprint drafts before submitting them to journals for peer review, and to their final drafts (postprints) after they have been peer-reviewed and accepted for publication by a journal.

Individual institutions cannot peer-review their own research output (that would be in-house vanity-publishing).

And global repositories like arxiv or pubmedcentral or citeseerx or google scholar cannot assume the peer-review functions of the thousands and thousands of journals that are actually doing the peer- review today. That would add billions to their costs (making each into one monstrous (generic?) megajournal: near impossible, practically, if it weren't also totally unnecessary -- and irrelevant to OA and its costs).

NGS: "Also, users have said again and again that they prefer discovery by subject, which will be possible for semantic docs in local repositories or better indexes (probably built through better collaborations), but not now."

Search should of course be central and subject-tagged, over a harvested central collection from the distributed local IRs, not local, IR by IR.

(My point was that central deposit is no longer necessary nor desirable, either for content-provision or for search. The optimal system is institutional deposit (mandated by institutions as well as funders) and then central harvesting for search.

NGS: "I agree that it would be great if local repositories were more used, and eventually, the systems will be in place to make it possible, but every study I've seen still shows local repository use to remain disappointingly low, although some universities are doing better than others."

"Use" is ambiguous, as it can refer both to author use (for deposit) and user use (for search and retrieval). We agree that the latter makes no sense: users search at the harvester level, not the IR level.

But for the former (low author "use," i.e., low levels of deposit), the solution is already known: Unmandated IRs (i.e., most of the existing c. 1500 IRs) are near empty (of OA's target content, which is preprints and postprints of peer-reviewed journal articles) whereas mandated IRs (c. 150, i.e.m 1%!) are capturing (or on the way to capturing) their full annual postprint output.

So the solution is mandates. And the locus of deposit for both institutional and funder mandates should be institutional, not central, so the two kinds of mandates converge rather than compete (requiring multiple deposit of the same paper).

For the special case of arxiv, with its long history of unmandated deposit, a university's IR could import its own remote arxiv deposits (or export its local deposits to arxiv) with software like SWORD, but eventually it is clear that institution-external deposit makes no sense:

Institutions are the universal providers of all peer-reviewed research, funded and unfunded, across all fields. One-stop/one-step local deposit (followed by automatic import. export. and harvesting to/ from whatever central services are needed) is the only sensible, scaleable and sustainable system, and also the one that is most conducive to the growth of universal OA deposit mandates from institutions, reinforced by funder mandates likewise requiring institutional deposit, rather than discouraged by gratuitously requiring institution-external deposit.

NGS: "Inter-institutional repositories by subject area (however broadly defined) simply work better, such as arXiv or even the Princeton-Stanford repository for working papers in the classics."

"Work better" for what? Deposit or search? You are conflating the locus of search (which should, of course, be cross-institutional) with the locus of deposit, which should be institutional, in order to accelerate institutional deposit mandates and in order to prevent discouraging adoption and compliance because of the prospect of having to deposit the same paper in more than one place.

(Yes, automatic import/export/harvesting software is indifferent to whether it is transferring from local IRs to central CRs or from central CRs to local IRs, but the logistics and pragmatics of deposit and deposit mandates -- since the institution is always the source of the content -- make it obvious that one-time deposit institutionally fits all output, systematically and tractably, whereas willy-nilly IR/CR deposit, depending on fields' prior deposit habits or funder preferences is a recipe for many more years of the confusion, inaction, absence of mandates, and near-absence of OA content that we have now.)

NGS: "Currently, universities are paying external middlemen an outsized fee for validation and packaging services. These services can and should be brought "in-house" (at least as an ideal/ goal to develop toward whenever the opportunities can be seized) except in cases where prices align with value, which occurs still with some society and commercial publications."

I completely agree that along with hosting their own peer-reviewed research output, and mandating its deposit in their own IRs, institutions can also use their IRs (along with specially developed software for this purpose) to showcase, manage, monitor, and measure their own research output. That is what OA metrics (local and global) will make possible.

But not till the problem of getting the content into OA IRs is solved. And the solution is institutional and funder mandates -- for institutional (not institution-external) deposit.

NGS: "To the extent that an arXiv or the inter-institutional repository for humanities research which will be showing up in 3-7 years moves toward offering these services, they are clearly preferable to old fashioned subscription models (since the financial support is for actual services) and current local repositories which do not offer everything needed in the value chain (as listed in Van de Sompel et al. 2004)."

(1) The reason 99% of IRs offer no value is that 99% of IRs are at least 85% empty. Only the 1% that are mandated are providing the full institutional OA content -- funded and unfunded, across all disciplines -- that all this depends on.

(2) The central collections, as noted, are indispensable for the services they provide, but that does not include locus of deposit and hosting: There, central deposit is counterproductive, a disservice.

(3) With local hosting of all their research output, plus central harvesting services, institutions can get all they need by way of search and metrics, partly through local statistics, partly from central ones.

NGS: " I remember when I first read an article quoting a researcher in an arXiv covered field who essentially said that journals in his field were just for vanity and advancement, since all the "action" was in arXiv (Ober et al. 2007 quoting Manuel 2001 quoting McGinty 1999) -- now think about the value of a repository that doesn't just store content and offer access."

This familiar slogan, often voiced by longstanding arxiv users, that "Journals are obsolete: They're only for tenure committees. We [researchers] only use the arxiv" is as false, empirically, as it is incoherent, logically: It is just another instance of the "Simon Says" phenomenon: (Pay attention to what Simon actually does, not to what he says.)

Although it is perfectly true that most arxiv users don't bother to consult journals any more -- using the OA version in arxiv only, and referring to the journal's canonical version-of-record only in citing -- it is equally (and far more relevantly) true that they all continue to submit all those papers to peer-reviewed journals, and to revise them according to the feedback from the referees, until they are accepted and published.

That is precisely the same thing that all other researchers are doing, including the vast majority that do not self-archive their peer-reviewed postprints (or, even more rarely, their unrefereed preprints) at all.

So journals are not just for vanity and advancement; they are for peer review. And arxiv users are just as dependent on that as all other researchers. (No one has ever done the experiment of trying to base all research usage on nothing but unrefereed preprints and spontaneous user feedback.)

So the only thing that is true in what "Simon says" is that when all papers are available, OA, as peer-reviewed final drafts (and sometimes also supplemented earlier by the prerefereeing drafts) there is no longer any need for users or authors to consult the journal's proprietary version of record. (They can just cite it, sight unseen.)

But what follows from that is that journals will eventually have to scale down to becoming just peer-review service-providers and certifiers (rather than continuing also to be access-providers or document producers, either on-paper or online).

Nothing follows from that about the value of repositories, except that they are useless if they do not contain the target content (at least after peer review, and, where possible and desired by authors, also before peer review).

Harnad, S. (1998/2000/2004) The invisible hand of peer review. Nature [online] (5 Nov. 1998), Exploit Interactive 5 (2000): and in Shatz, B. (2004) (ed.) Peer Review: A Critical Inquiry. Rowland & Littlefield. Pp. 235-242.

NGS: "Do I think the financial backing will remain in place? It depends on the services actually offered and to what extent subject repositories could replace a patchwork system of single titles offered by a patchwork of publishers."

At the moment the issue is whether arxiv, such as it is (a central locus for institution-external deposit of institutional research content in some fields, mostly physics, plus a search and alerting service), can be sustained by voluntary sub-sidy/scription -- not whether, if arxiv also somehow "took over" the function of journals (peer review), that too could be paid for by voluntary sub-sidy/ scription...

NGS: "Universities could save a great deal by refusing to pay the same overhead over and over again to maintain complete collections in single subject areas (not to mention paying for other people's profits)."

I can't quite follow this: You mean universities can cancel journal subscriptions? How do those universities' users then get access to those cancelled journals' contents, unless they are all being systematically made OA? Apart from those areas of physics where it has already been happening since 1991, that isn't going to happen in most other fields till OA is mandated by the universal providers of that content, the universities (reinforced by mandates from their funders).

Then (but only then) can universities cancel their journal subscriptions and use (part of) their windfall saving to pay (journals!) for the peer-review of their own research output, article by article (instead of buying in other universities' output, journal by journal).

NGS: "More importantly, more could be done to make articles useful and discoverable in a collaborative environment, from metadata to preservation, so that the value chain is extended and improved (my sci-fi includes semantic docs, not just cataloged texts, and improved, or multi-stage, peer review, or peer review on top of a working papers repository)."

All fine, and desirable -- but not until all the OA content is being provided, and (outside of physics), it isn't being provided -- except when mandated...

So let's not build castles in Spain before we have their contents safely in hand.

NGS: "I think there's been plenty of 'chatter' to indicate that the basic assumptions in conversations between universities are changing (see recent conference agendas), so that we can expect to see more and more practical plans to collaborate on metadata, preservation, and , yes, publications."

I'll believe the "chatter" when it has been cashed into action (deposit mandates). Till then it's just distraction and time-wasting.

NGS: "My head spins to think of the amount of money to be saved on the development of more shared platforms, although, the money will only be saved if other expenditures are slowly turned off."

All this talk about money, while the target content -- which could be provided at no cost -- is still not being provided (or mandated)...

NGS: "Sandy mentioned in another post that she [he] would hope for arXiv like support for university monographs..."

Monographs (not even a clearcut case, like peer-reviewed articles, which are all, already, author give-aways, written only for usage and impact) are moot, while not even peer-reviewed articles are being deposited, or mandated...

NGS: "Open access and NFP publications which do offer the full value chain have been proven to have much lower production costs per page than FP publishers and they do not suffer any impact disadvantages -- and these are still operated on a largely stand-alone basis, without the advantages that can be gained by sharing overhead."

Cash castles in Spain again, while the free content is not yet being provided or mandated...

NGS: "Maybe local repositories really are the way to go, since then each institution has more control over its own contribution, but the collaboration and the support will still need to occur to support discovery (implying metadata, both in production and development of standards and tools) and preservation."

No, search and preservation are not the problem: content is.

NGS: "I suppose another problem with local repositories, however, is that a consensus is far less likely to unite around local repositories as a practical option at this juncture -- the case can't just be made with words, you need the numbers and arXiv has them -- and while I am interested to see strong local repositories emerge, there is greater sense in supporting what can be achieved, since we need more steps in the right direction."

"The numbers" say the following:

Physicists have been depositing their preprints and postprints spontaneously (unmandated) in arxiv since 1991, but in the ensuing 20 years this commendable practice has not been taken up by other disciplines. The numbers, in other words, are static, and stagnant. The only cases in which they have grown are those where deposit was mandated (by institutions and funders).

And for that, it no longer makes sense (indeed it goes contrary to sense) to deposit them institutional-externally, instead of mandating institutional deposit and then harvesting centrally.

And the virtue of that is that it distributes the costs of managing deposits sustainably, by offloading them onto each institution, for its own output, instead of depending on voluntary institutional sub-sidy/scription for obsolete and unnecessary central deposit.

(See also the "denominator fallacy," which arises when you compare the size of size of central repositories with the size of institutional repositories: The world's 25,000 peer-reviewed journals publish about 2.5 million articles annually, across all fields. A repository's success rate is the proportion of its annual target contents that are being deposited annually. For an institution, the denominator is its own total annual peer-reviewed journal article output across all fields. For a central repository, it is the total annual article output -- in the field(s) it covers -- from all the institutions in the world. Of course the central repository's numerator is greater than any single institutional repository's numerator. But its denominator is far greater still. Arxiv has famously been doing extremely well for certain areas of physics, unmandated, for two decades. But in other areas arxiv is not not doing so well, relative to the field's true denominator; and most other central repositories are likewise not doing well, In fact, it is pretty certain that -- apart from physics, with its 2-decade tradition of deposit, plus a few other fields such as economics (preprints) and computer science -- unmandated central repositories are doing exactly as badly unmandated institutional repositories are doing, namely, about 15%.)

Stevan Harnad
American Scientist Open Access Forum

Belgium's 4th Green OA Mandate, UK's 29th, Planet's 140-141st

University of Ghentand University of Reading

---

Please register your own university's mandate in ROARMAP too, to track progress and to encourage other universities to adopt mandates of their own.

Congratulations to MIT for this extremely helpful streamlining of the deposit process:

"MIT Libraries began to investigate how SWORD and SWAP could facilitate external contributions by publishers... Entering long and complex information about articles is avoided with the MIT Libraries’ customized submission interface.  Only two pieces of metadata are required for already published papers: the name of the authorizing MIT author and a DOI or URL. If the paper is unpublished, four fields are requested."

Although entering metadata is not really that complicated and time-consuming at all, we know it is difficult to persuade those who have never deposited a paper in an institutional repository of this fact.

So reducing deposit to just entering a name and URL would be a huge step forward in facilitating mandate compliance -- and of course also in encouraging unmandated deposit.

I hope we will implement this quickly for EPrints repositories too.

I am, however, far less sanguine about the second -- publisher-deposit -- option, especially for mandated deposit:

'the use of SWORD and SWAP with the DSpace repository at MIT is part of a larger strategy to improve collaboration with publishers, facilitating a “push” of large amounts of content into a repository without necessitating a platform-specific solution. Ultimately this “publisher template” could be used with other repository platforms such as Fedora and EPrints. Richard Rodgers, Head of Software Development at MIT Libraries, says, “If we do this right there will be no code to share. SWORD and SWAP are already open and accessible. We have localized their use to accommodate MIT-specific metadata.”

It might be alright to quietly provide a way for publishers to facilitate IR deposit, but it would be a huge strategic error to give them an active or essential hand in it.

All the power of self-archiving (and of self-archiving mandates from institutions and funders) comes from the fact that it is the author and the author's institution (and funder) that does it, mandates it, and monitors compliance.

Self-archiving -- its doing and its timing -- is all in the research community's own hands. Publisher deposit is not.

The little extra content that publisher-deposit or publisher-facilitated deposit might add does not counterbalance the additional author confusion, deposit delay, diffusion of responsibility and difficulty in compliance-monitoring that it is likely to introduce into institutional mandates, as it has already done with those funder mandates that allow fundees to offload their mandate fulfillment obligations onto publishers. The problem is especially with specifying and monitoring the fulfillment conditions for deposit mandate compliance. (We always have to remember that publishers are neither employees nor fundees, and hence they are not the ones subject to the deposit mandates).

(What kind of mandate is it if it says "You must deposit -- unless your publisher does it for you..." How is it even to be monitored whether and when the mandate has been complied with?)

So if repositories implement some sort of back door for publisher-facilitated deposit, it is important to keep a low profile on it and to stress that on no account should it be stipulated or relied on as one of the ways to fulfill a deposit mandate: Complying with the mandate must be entirely the responsibility of the author, and the monitoring and verification of compliance must be based entirely on steps taken by the author, not steps the authors leave to a publisher to (possibly) take (sometime) on their behalf...

Stevan Harnad
American Scientist Open Access Forum

Professor Steven Hyman, Provost of Harvard, the first US University to mandate Open Access, has submitted such a spot-on, point for point response to President Obama’s Request for Information on Public Access Policy that if his words are heeded, the beneficiaries will not only be US research progress and the US tax-paying public, by whom US research is funded and for whose benefit it is conducted, but research progress and its public benefits planet-wide, as US policy is globally reciprocated.

Reproduced below are just a few of the highlights of Professor Hyman’s response. Every one of the highlights has a special salience, and attests to the minute attention and keen insight into the subtle details of Open Access that went into the preparation of this invaluable set of recommendations.

[Hash-marks (#) indicate three extremely minor points on which the response could be ever so slightly clarified -- see end.]

“The public access policy should (1) be mandatory, not voluntary, (2) use the shortest practical embargo period, no longer than six months, (3) apply to the final version of the author’s peer-reviewed manuscript, as opposed to the published version, unless the publisher consents to provide public access to the published version, (4) [# require deposit of the manuscript in a suitable open repository #] immediately upon acceptance for publication, where it would remain “dark” until the embargo period expired, and (5) avoid copyright problems by [## requiring federal grantees, when publishing articles based on federally funded research, to retain the right to give the relevant agency a non-exclusive license to distribute a public-access copy of his or her peer-reviewed manuscript ##]…

“If publishers believe they cannot afford to allow copies of their articles to be released under a public-access policy, they need not publish federally funded researchers. To date, however, it appears that no publishers have made that decision in response to the NIH policy. Hence, federally funded authors remain free to submit their work to the journals of their choice. Moreover, public access gives authors a much larger audience and much greater impact…

“If the United States extends a public-access mandate across the federal government, then lay citizens with no interest in reading this literature for themselves will benefit indirectly because researchers will benefit directly…. That is the primary problem for which public access is the solution…

“It doesn’t matter whether many lay readers, or few, are able to read peer-reviewed research literature or have reason to do so. But even if there are many, the primary beneficiaries of a public-access policy will be professional researchers, who constitute the intended audience for this literature and who depend on access to it for their own work….

“Among the metrics for measuring success, I can propose these: the compliance rate (how many articles that the policy intends to open up have actually been opened up); the number of downloads from the public-access repositories; and the number of citations to the public-access articles. As we use different metrics, we must accept that [### we will never have an adequate control group: a set of articles on similar topics, of similar quality, for which there is no public access ###]….

---

Three suggestions for clarifying the minor points indicated by the hash-marks (#):

[#”require deposit of the manuscript in a suitable open repository” #]

(add: “preferably the fundee’s own institutional repository”)

[##”requiring federal grantees, when publishing articles based on federally funded research, to retain the right to give the relevant agency a non-exclusive license to distribute a public-access copy of his or her peer-reviewed manuscript” ##]

(add: “the rights retention and license are desirable and welcome, but not necessary if the publisher already endorses making the deposit publicly accessible immediately, or after the allowable embargo period”)

[### "we will never have an adequate control group [for measuring the mandate's success]: a set of articles on similar topics, of similar quality, for which there is no public access" ###]

(add: “but closed-access articles published in the same journal and year as mandatorily open-access articles do provide an approximate matched control baseline for comparison”)


Stevan Harnad
American Scientist Open Access Forum

Cornell University Library has proposed a "Collaborative Business Model" for funding the worldwide Physics ArXiv that it hosts (see white paper).

"arXiv will remain free for readers and submitters, but the Library has established a voluntary, collaborative business model to engage institutions that benefit most from arXiv."

Here's an alternative to this voluntary institutional sub-sidy/scription model whose sustainablity -- through all economic times, tough and tender -- is less founded on blind faith:

Institutions have many self-interested reasons for wanting to host, archive, manage, monitor, measure and showcase their own research article outputs. The annual scale of their own local article output is also manageable and sustainable at the institutional level, within each institution's existing infrastructure:

Carr, L. The Value that Repositories Add
Swan, A. The Business of Digital Repositories
Harnad, S. Institutional vs. Central Repositories

Hence what will happen is that instead of trying to sustain a central repository like Arxiv -- most of whose costliness derives from the fact that it is a single direct locus of deposit and archiving from all institutions, worldwide -- direct deposit and hosting (and its costs) will instead be offloaded and distributed across the network of institutional repositories, with Arxiv becoming merely another central harvester, providing global search services (sustainable if it provides functionality that can compete with other OAI services or Google Scholar).

But voluntary sub-sidy/scription will no doubt sustain things for a while. (Things do seem to catch on rather slowly in this domain...)

Stevan Harnad
American Scientist Open Access Forum