Intota, Work package 5: Workflows, Workpackage 4: APIs

December 6, 2012 Amy Devenney

Patron-Driven Acquisition (PDA) is a method of purchasing materials for a library based on a known patron demand.[1] It is an example of the ‘Just-in-time’ acquisitions model which is a reaction by the library to ensure the needs of the user are met, as opposed to the ‘Just-in-case’ acquisitions model which is the traditional way in which materials are selected by a Librarian in anticipation of the user’s needs. However PDA is not a new idea, it has been around for about ten years, longer if you include Inter-Library Loans (ILL)! Although ILL’s do not inform purchasing decisions and are only a loan of an item it does allow patrons to decide on which resources a budget is spent on. This concept of allowing patrons to decide on the items that are permanently added to the collection is becoming increasingly popular with libraries to ensure that their budgets are spent on items that satisfy the needs and demands of the patrons. The rise in the implementation of PDA by libraries means that any of the new next generation library and web-scale management system must have the capacity to work with this acquisition model, among others, and as such will be a useful criterion for us in the evaluation of the suitability of Intota of the Huddersfield and the UK Higher Education marketplace.      

As PDA can be used for both print and electronic items there are a number of different methods of implementing PDA. One method is the selection of items for purchase by subject specialists (lecturers/professionals/researchers) within the University. The selection can occur in a number of different ways, at the Brigham Young University subject specialists fill in a form on the library catalogue with the details of the item they would like to have purchased.[2] This then goes directly to the acquisitions team who order the item without consulting with the Librarian. A similar but more automated process is currently being developed at Huddersfield with the MyReading software. This in-house software is being developed to allow lecturers to upload their reading lists (identifying essential, recommended and additional reading) on to a VLE and be viewable by students. These reading lists working with ASIS (the Applicant and Student Information System which holds all the academic and personal data for applicants and students including the modules the students are enrolled on) will then inform acquisition decisions by having a formula within the system to automatically create orders based on the categorisation (essential, recommended or additional) of the book by the tutor, the number of students on the module and the format of the items. For example, a formula may be:

If an ebook is marked as essential reading, buy 1 copy for every 25 students on the module but for a print copy only buy 1 copy for every 10 students.

These automatic orders will then either be routed directly to the acquisitions team (if essential) or to the Librarian (if recommended or additional) for a decision on purchase to be made.

Similarly many libraries use patron suggestions to inform purchases made by the library.[3] This is where any patron can suggest an item that they think would be advantageous for the library to purchase; it could be by an electronic form, a paper form or even a verbal recommendation. The Librarian then considers the request before passing the details on to the acquisitions team if the decision is to purchase the item. 

E-book PDA is another popular and easy method of PDA. The Librarians will decide on which subject areas/Dewey number that they would like to be available for PDA and will then up-load all the relevant e-book catalogue records from the supplier on to the catalogue. Trigger points for purchase will also be chosen by the Librarian, for example, in a PDA trial Huddersfield it was decided that a second five minute preview of an item would trigger the purchase. However, there are a number of different triggers including the number of users, time spent with the item, number of pages viewed, etc. This allows all the e-books to be searchable by the patrons and if the trigger point is met a purchase will automatically occur. 

Another PDA model employed at Huddersfield is Hey Presto which is an order generated in response to multiple holds on an item. Hey Presto was developed in-house to send an alert to the relevant subject team when the number of holds on an item exceeded the number of copies held by the library. This informs the subject teams of where the demand for a book, often in response to a recommendation by a tutor, is not being sufficiently met by the library. After receiving the alert the team can then make a decision on how many copies to buy and the format in which to buy the item. 

One form of PDA that Huddersfield does not currently use is ILL purchase on demand. This is where after having received an Inter-Library Loan request the team then decides whether to fulfil the loan request or whether to purchase the item for the collection permanently. Another way of looking at this is to call it print PDA!

The majority of the studies on PDA have tended to be favourable showing that items purchased by PDA are more cost effective than those purchased under the normal selection method because they generally have higher circulation.  At University of West-Maddison 73% of items bought in response to an ILL request circulated twice or more in a two year period as opposed to 6% of the items acquired through the normal selection method.[4] At Brigham Young University (BYU) the e-book PDA gave similar results as they were 26% cheaper than the ones obtained through the traditional methods and were used 13.75 times more. [5] The same study at BYU also demonstrated that although many of the print books obtained through PDA were the same price as those purchased by the traditional selection method they were circulated more frequently giving a lower cost per use (CPU).[6] Statistics from the University of Huddersfield e-book PDA in 2011 showed that PDA titles had double the usage of a non-PDA title. The average number of views per month for a non-PDA title was 0.966 as opposed to the 2.03 views of a PDA title. These studies and many others have shown that PDA is a very successful and cost-effective way of acquiring items for the library and satisfying patron demand, and is an acquisitions model which should be seriously considered by libraries as they face decreasing budgets.

Despite this some Librarians still harbour reservations about the use of PDA in academic libraries. One of their main concerns about this acquisition model is that it may result in the collection becoming less academic and more ‘popular’, while another concern is that because the budgets for these models of acquisitions are often shared and there is no easy way of monitoring the spend one subject area could spend more than their percentage of the budget to the disadvantage of another.[7] However, recent research has demonstrated that these fears are unfounded. An investigation by Lisa Shen et al. investigated the difference in the academic quality of books chosen by PDA and those chosen by the librarians. By asking librarians to choose the items they would order from a list of books and then presenting the same list to patrons they researchers found that the selections were very similar in content, thus disagreeing with the notion that the items selected by the patrons would not be as academic as those chosen by the librarians.[8] The other concern of the librarians over PDA concerning the inability to ensure that the subject’s quota of the budget is correctly spent is being looked at by e-book providers. Some e-book providers have developed software that allows you to establish separate budgets on the e-book platform and assign dewy numbers to these budgets so that when a book is bought it will then deduct the money from the relevant budget. This software can also send an alert to the correct subject team when the budget is nearly spent to ensure you either withdraw you titles from the catalogue, making them unavailable via PDA, or you replenish your account. Another concern, which is unique to e-book PDA, is the uploading and subsequent deleting of numerous catalogue records to the library catalogue which can take time and is not instantaneous. It took the University of Huddersfield 3-4 days to upload 120,000 records and then at least a day to delete them. The main concern with this delay is that it is not possible to instantly delete them if the budget is spent; however with the implementation of Intota it is possible that this problem may be resolved. Rather than uploading the records to the catalogue and then deleting them, the books will be made available through the Knowledge Base by switching them on and then to remove them you switch them off.

Growing evidence is showing that PDA is a very successful acquisitions model which allows the library to increase its holdings with cost effective specialist academic content that is wanted by the patrons and has a successful circulation rate. Although librarians have some concerns over these methods recent research appears to proving them unfounded. Therefore it is probable that this method of acquisition will prevail and perhaps pervade over the traditional ‘just-in-case’ model of selection. It is crucial then that a new next generation library and web-scale management system such as Intota must have this acquisitions model built into the workflow.