The PI Warrior and The Bag of Gold: OA Metadata & University Accounting Practice

Peter Burnhill, EDINA (Jisc centre for service development & delivery), University of Edinburgh


The warrior I have in mind as hero for this story is the Principal Investigator (PI) who plays a vital role in leadership and decision-making within a university and with research funders. This role includes research reporting and the formation of cross-institutional research teams, with the wish for recognition, impact and future funding.

The brief is to provide a Use Case for ‘Identifying common and required transaction points where information (and sometimes money) is exchanged’. So central to the story is a workflow during which key information for the record of Open Access publication and research reporting can be generated as metadata. Relevant metadata fields are indicated in [square] brackets – and I invite others to similarly annotate according to CERIF, RIOXX and the like. An earlier version was made available to Jisc as a briefing note, ‘The Pedestrian View”, from the RepNet project, so for this blog post I thought I should spice-up what otherwise risked being another unexciting title.

A key role in the story is played by the cost accountant’s Purchase Order and the subsequent handling of the Invoice. The purpose of the story is to try to gain leverage from the transactions that take place about money in order to brings together key metadata. This should include not only text fields such as names and titles, but also the numeric identifiers that assist unambiguous linkage, such as the [Project ID], the [Funder ID], the [ISSN] and (eventually) the [DOI]. Done properly, it could also support the use of the [ISNI] for both the publisher and the university, and may also cater for the [ISNI] or [ORCID] for the author(s). This might also be the occasion for the purchaser to specify the type of OA to be purchased through the payment of Gold (CC-BY, say). Whether payment is made before or after actual delivery of the service in question (that the article is available as Open Access on the terms agreed) is an open policy question.

The PI and the Author(s) are two main characters in this use case, alongside that of the Publisher (ISNI] from whom the service of Open Access is bought with a bag of Gold. The Funder who provides the PI and the Institution [ISNI] with the pot of Gold is off-stage for the scene that is presented.

The Head of Department (HoD) could have more than a walk-on part especially if the pot of Gold is centrally managed in order that the use of this Gold is for research output that occurs beyond the formal award period. That appearance would trigger thoughts about the policy with Departments (or some other academic planning unit) for assisting Gold purchases for the output from ‘unfunded’ research (project) activity from any special funding pots set up within the Institution, and so is left for a scene later in the programme.

Additional assumptions to set the scene

a) More generally, this use case takes it as a given that university finance offices require (or, at the very least, prefer) to make payment on presentation of an Invoice only against a previously issued Purchase Order (PO) [POid?] or other formal pre-notification. That Purchase Order (PO), which specifies the product/service to be purchased and which includes note of the amount and the [account number] to be charged, is sent to the supplier of the product/service. There is implicit promise that payment will be made on presentation of an invoice by the supplier [INVid?], given satisfaction that the product/service is delivered.

b) The use of Gold to buy Open Access for that journal article should be treated like much any other purchase using ‘restricted funds’ at a university: someone needs to make the decision on what to buy and there needs to be an agreed financial process that leaves some audit trail. In this instance we also would like to have some understanding about how there is some generation of metadata that can be used for a variety of purposes.

c) Decision on which article should be bought with Gold involves academic judgement, so we want academics involved – even if the academic support services (typically the Library) and the grants and the finance offices provide the infrastructure for that decision making. The candidates are the PI and the Author(s), and maybe the HoD.

d) PIs have designated authority for many decisions. Support for this warrior class is central for the successful implementation of a university’s research strategy, most obviously in generating for research assessment exercises, reputation and the flow of ‘unrestricted funds’ as contributions to the full economic cost (fEC) in an institution. However, typically, the PI knows that to be successful s/he has to obey the rules set by the institution’s finance office in order to make payments from a given project grant.

The play begins

Prologue. This is story of a PI who leads a team of researchers, drawn from more than one institution [ISNI], to conduct a grant-funded research project [FunderID, ProjectID]. The plot revolves around the research output from that project, a journal article [DOI], recognising that the DOI is assigned only towards the end of the scene. Authorship may often include the PI, but for the sake of the play it need not, and in any event the PI may not always be the first/corresponding author who has contact with the journal publisher. What is important is that the article is recognised to have more than one author, employed at more than one institution. An even simpler but ultimately a very parochial and unhelpful story would be that of a single PI who is also the sole Author.

i) Enter Dr Jan Arrowsmith, at the University of Mercia [ISNI}, who is the Principal Investigator (PI) for a research grant from [FunderID]. There is agreement that funds in that grant may be used as Gold to make research publications available under terms of OA, using CC-BY (say). There are also funds held centrally at the University of Mercia for similar purposes. Dr Arrowsmith leads a team of five researchers, two of whom are employed by the University of Wessex, including Harry Crosbie in her department and Dr Tim Brown, based at the University of Wessex.

ii) Separately, Harry and Tim are corresponding with representatives of the Publishers [ISNIs] of Journals X and Y [ISSNs] in their respective roles as First (Corresponding) Author of two articles that have been submitted and (provisionally) accepted for publication. There are several co-authors for each, including those in the project employed at both universities, and an (unfunded) academic at the University of Cornwall. In both cases there is opportunity to make the journal article Open Access in return for a fee (Gold OA).

a) Harry (Author) speaks to Jan (PI) about whether this is an appropriate use of funds: they agree that it is.

b) There is similar discussion between Tim (Author) and Jan (PI).

iii) Dr Arrowsmith (or typically her Administrator) raises a Purchase Order (PO) within the University of Mercia, for both (a) and (b). What follows is generic to Harry and Tim.

iv) The PO specifies the service that is to be purchased (the terms of OA required / on offer), and this could be linked to a metadata record that specifies [FunderID] and [ProjectID], together which whatever bibliographic detail is available at that moment, typically only the [ISSN] is certain at this point.

  • typically the issue/volume and the DOI are unlikely to be fixed at this moment, and even the title may still be a bit hazy.

v) The Publisher receives the PO, with the specification explicit, including the [FunderID] and the [ProjectID], and regards this as sufficient assurance with which to process the article in a way that would recognise this as an OA article and against which an Invoice could be raised with high likelihood of prompt and full payment.

vi) Once the Publisher [ISNI] has arranged publication in an issue of the journal [ISSN] the full bibliographic information for the article is determined: this includes the assignment of the [DOI], This is opportunity to link these with [FunderID] and the [ProjectID] in a metadata record as well as include them in the full text of the article. The Publisher is then in a position to make the article [DOI] available under the specified terms of OA, according to the PO. (There is also the prospect that the DOI might contain some indicator that the publisher’s version of this article has indeed been made OA.)

vii) The Publisher, having carried out the service that was purchased then presents the Invoice to the purchaser (Dr Arrowsmith the PI) for payment, copying in the Author, together with the metadata record, marked “Please give this to your CRIS …”. Both the [FunderID] pointer to the full bibliographic information [including DOI] would allow successful delivery of the service to be checked – just as is done for all purchases in universities – and perhaps the safe deposit of the article in question.

viii) As with any such purchase from her grant [ProjectID], Dr Arrowsmith the PI (or typically her Administrator) arranges payment with the Finance Office system at the University of Murcia; Dr Arrowsmith knows that this works simply and without fuss whenever there is a Purchase Order against which to match an Invoice.

  • There is also now prospect of using that general Gold OA fund that Dr Arrowsmith found out about when she attended a recent Library Committee. Although she should have used a particular account code in the Purchase Order for the two articles in question, she discovers to her delight that this will pay for the Gold OA access for those articles that she published after her previous grant project ended. (In fact she discovers that in future she can apply to use this central fund and will not have to spend from her own research grant – and so attendance at that conference in the US can be now be afforded.)

ix) There is happy outcome to this story associated with tracking the payment of Gold from within the seemingly dull workflow of the university research and finance offices.

a) PI and the University of Murcia (via CRIS, IR or some system) have sufficient metadata for many of the things that they need to do for research reporting, both to notify the Funder [FunderID] and [ProjectID], that preparation for the REF and for the annual report by the University, supplying all required information and with [DOI] pointer to the full bibliographic information.

b) The Library (and each shared service facility) also has what it needs to ensure custody and ‘show case’ of the full text.

c) The Research Office, and the Library, are assured that PI (Dr Arrowsmith), a key member of the University’s warrior class, expresses delight that all her project team are made happy: the Author, Dr Brown has what he needs to ensure that the IR/CRIS at the University of Wessex are notified that the article has been deposited as per mandate and is now available under OA, as CC-BY say; the same was true for that (unfunded) academic at the University of Cornwall.

d) Now that this research output is Open Access there is prospect of greater impact for research output, with greater prospect of future funding and enlightenment for a wider spectrum of researchers and students.

The PI Warrior and The Bag of Gold: OA Metadata & University Accounting Practice

7 Responses

  1. This is a good summary of the process flows involved in the Gold OA model, and the opportunities to capture metadata in the process. You have rightly alluded to the additional challenges involved where the pot of Gold is centrally managed, and the need for a formal policy and decision-making process to govern access to this, and think this is the most pressing challenge for many academic institutions at present. I agree any such process must be academically led, but this of course brings with it a consequent increase in the administrative burden placed on Heads of Department (or whoever else takes on the decision-making role), and an erosion of the autonomy and authority of the PI warrior.

    The other challenge is the scalability of the process. While a standard PO, receipting and invoice payment process is manageable at low transaction volumes, the additional administrative burden entailed as the number of APCs processed runs into the hundreds and thousands becomes significant. Publisher prepayment arrangements and intermediaries can potentially alleviate some of this, but most have yet to establish robust and streamlined mechanisms to allow institutions to approve expenditure before it is incurred.

    Rob Johnson May 1, 2013 at 6:59 pm #
    • Thanks Rob. As you say, two big challenges:

      a) academically-led decision making for centrally-managed funds.
      b) scalability ‘as the number of APCs processed runs into the hundreds and thousands’
      - for proof-of-delivery and payment (in whatever order)
      - for the supporting decisions & placement of orders.

      My purpose was to view this through the eyes of not only the PI warrior (for the research grant) but also for the Institutional Finance/Grants Office who will have to be on board sooner or later. Maybe time for Harry (or Tim) to tell the story from their pov – complete with credit card and discussion with the gatekeeper of the Central Pot of Gold

      Peter Burnhill May 3, 2013 at 12:45 pm #
  2. I’ve been sent some questions/comments directly resulting from this blog post to which I respond here, one by one:

    Q: Gold OA intermediaries will play a significant role in the future in this space, similar to the role subscription agents play now. A number of institutions are piloting the Jisc APC system,

    Reply: Intermediaries can always play a useful role, but usually these have operated between institutions (typically their libraries) and publishers for journals. There is much greater granularity with articles for which it is authors & PIs (or Heads of Departments) that have decisions to make. Time for a another story … (as use case)?

    Peter Burnhill May 2, 2013 at 10:16 am #
  3. Another two comments made to me directly:

    Q1: A Purchase Order is very rarely raised by researchers. It depends on how well drilled/efficient the local department admin is. Generally raising POs add time to the overall transaction which is not ‘a good thing’.

    As a PI/author I confess that I don’t get to do the Purchase Orders myself , but I do know that our administrators tell me that their use speeds up the overall transaction – at least in terms of ensuring that the supplier gets paid and does not hassle and cause the author/researcher (or administrator) grief. Also means that the PI has some budgetary control. Of course if you pay in advance and don’t attend to whether the thing (Gold OA) you have bought actually exists (as Open Access – ie downloadable outside your university IP range) and preferably is collected in your national/institutional facility …

    Q2: A small but important number of Gold OA transactions are facilitated by the use of researchers own credit cards. We are trying to iron this out although many publishers offer this option.

    To my mind, the fact of Q2 argues in favour of the use of a PO and is why researchers/authors need the support of institutional processes. I’m not given to being a ‘process man’ but there is high risk of chaos – as well as all those fake OA journals taking authors for a ride …

    Peter Burnhill May 3, 2013 at 12:30 pm #
  4. Good coverage of the issues around publishing articles in APC-financed journals, but what I am missing here is thoughts on how
    (1) journals not charging per-article APCs (e.g. no APCs, or institutional membership, or researcher membership)
    (2) publications other than classical articles (e.g. data or software)
    fit in.

    Daniel Mietchen June 4, 2013 at 12:24 am #
  1. The PI Warrior and The Bag of Gold: OA Metadata... - May 1, 2013

    [...] The brief is to provide a Use Case for ‘Identifying common and required transaction points where information (and sometimes money) is exchanged’. So central to the story is a workflow during which key information for the record of Open Access publication and research reporting can be generated as metadata. Relevant metadata fields are indicated in [square] brackets – and I invite others to similarly annotate according to CERIF, RIOXX and the like. An earlier version was made available to Jisc as a briefing note, ‘The Pedestrian View”, from the RepNet project, so for this blog post I thought I should spice-up what otherwise risked being another unexciting title.  [...]

Leave a Reply

six − 3 =