In November 1994 members of the library staff met to discuss the feasibility of setting up an experiment in Electronic Document Delivery (EDD). This was a result of the continued increase of periodical prices above the rate of inflation; the increasing difficulty in subscribing to core journals caused by price increases and the need to create more space in the Library. The Deputy Librarian produced statistics which showed that 102 journals subscribed to by the Library cost over 800 pounds each i.e. 5 percent of the titles taken by the Library cost 30 percent of the total periodicals expenditure. Most of these were in the sciences and technology area.
The short term aim of the experiment was to see if EDD could save money on expensive journal subscriptions. The long term aim was to increase the cost effectiveness of obtaining journal articles, thus ensuring value for money, and to test the acceptability of a service delivering full text articles to users. At no point did we consider replacing existing document delivery services with electronic document delivery. The two would be run as complementary services.
A total of 4,000 pounds was made available. Using the Blackwell's Uncover Service the experiment was planned to run for six months maximum or until the money ran out. It was agreed to encourage the Library representatives from the University's Science Departments to actively participate in the experiment.
The 'Uncover' database was chosen because its journal title coverage of approximately 16,000 journals including good academic coverage, seemed extremely good. Of the 102 titles mentioned previously Uncover could supply 100. The average price of the faxed articles worked out at approximately 10.50 dollars although copyright charges did depend on the price of the actual journal i.e. the more expensive the journal the more expensive the copyright charge. The University of Bath Library paid for a password account, in which a designated sum was deposited. Blackwells debited this account monthly sending regular statements itemising each transaction. Advantages included increased security and a reduction of the copy charge for each article by 2 dollars.
The first trial started on March 1 1995 - mid May 1995
- 2 Departments, Maths and Chemistry, were selected to take part in the experiment, and a steering party of two library representatives and four library staff was established.
- Seven journal titles were selected; some which had previously been taken by the Library and cancelled; others were titles that the departments would like to subscribe to.
- The total annual subscription for these journals would be 8,000 pounds.
- All University staff and researchers at the University of Bath could order articles from any of these seven journals.
The above meant that we should be able to achieve our objective of assessing the viability of EDD in terms of quality and speed of delivery, and how such a service should be monitored and offered.
It would not give us a direct correlation between expensive current subscriptions and quantifiable measures of use, but would enable us to make comparisons to present subscriptions which may be of use if expensive journals needed to be cancelled in the future.
The trial was advertised in Staff and Student newsletters, as well as the combined Library and Computing Services newsletter. Fliers were put in all lecturers' and postgraduates' pigeon holes as well as the Library's noticeboards in the Departments.
Applications for EDD were originally only accepted by email, but eventually extended to receipt by post or in person. Guidelines were given on the fliers detailing what information should be included, and a template was provided on the web page. Web and email requests were sent to a special email address. In order to monitor cost and use, requests for articles via EDD were processed by two members of Academic Services who staffed the Reference and Information Point . All applications were first checked against the Library catalogue and holdings and if the item was not in stock, the request was processed and sent to the Uncover service. If the applicant gave a fax number, the request to Uncover specified that number and the article was faxed directly to the applicant. If not, requests were directed to the Library fax and then forwarded by internal mail.
After three months the original trial was reviewed. Take up of the service had been extremely disappointing i.e. only 9 requests. We felt that EDD could still be a useful service and to test out the system it was decided to continue with the project, but in a revised format from June 1995. The whole trial would again be reviewed in the Autumn.
1 June 1995 - 31 October 1995
- The service would continue to be limited to researchers and all university staff.
- They would be eligible to request any urgently required article (i.e. required within 24-48 hours) on the Uncover database that was not subscribed to by the Library.
- The experiment would last until the money ran out, unless the demand became overwhelming, in which case the service would be withdrawn for further consideration.
Although this trial would be carried out during the summer vacation, it was felt to be a good time as many of the lecturers caught up on their research during the summer; many postgraduates would be writing up and needing articles urgently, plus the new intake of postgraduates would be starting their projects and requiring articles.
Methodology continued as for the first trial with a few minor changes such as processing requests being limited to twice a day i.e. early morning and late afternoon. These changes took place because:
a) the increase in the number of requests meant processing and checking became quite time consuming which had implications on the rest of their workload and
b) the network was very busy between 11a.m. and 4 p.m. making the system difficult to access and very slow to search and retrieve article citations.
In order that we received feedback on the service, users were requested to return the front cover of the fax back to the Library with comments about the service, such as speed of delivery, quality of fax, whether the service was useful or not. Library staff also started noting the time spent on individual searches and how many orders were or were not successful.
Most of the feedback was made in the first few weeks of the new trial starting. After the first request many people did not bother to return the front sheet, and we only heard from them if there was a problem.
On the whole the service worked well and was felt to be a success. Statistics showed that 79 percent of requests made were fulfilled; 16 percent were unable to be fulfilled by Uncover and 5 percent of requests were already available in the Library. Comments such as "excellent service", "fast and efficient service", "excellent additional service complementing existing facilities" were made. The majority of take-up had been by teaching staff from nearly all of the Departments within the University. There was no evidence of non-academic members of staff using the service.
It was particularly successful because requests could be made electronically without users having to come in to the Library and filling in a form. Users increasingly wanted to order material via their PC and have it delivered to their desk!
Library staff working on the project felt that the back-up from Blackwells Help Desk had been excellent. Emails were responded to promptly, and if necessary articles were re-sent on the same day.
There were however some problems with the service:-
- As mentioned previously, actually accessing and searching the database, as well as inputting the information at certain times of the day became virtually impossible due to network delays. 10 minutes became the average time spent by Library staff dealing with each request, a cost of 2.50 pounds per request. Some requests took longer because of the system being slow, or because articles were not in the system and required double and/or triple checking via author, keyword or journal title. A list of journals available on Uncover would have prevented some of this unnecessary searching for titles not indexed.
- Although 16,000 journals were indexed by Uncover, some of the lecturers felt that coverage was poor. Often it was found that very current articles were also unavailable.
- The cost of the service was higher than expected with the average price of articles costing 18 dollars each.
- Library staff found the system rather "user unfriendly" with a lot of repetitive keying of passwords and commands being required. The lack of complete alphabetic order in author search displays was also unhelpful. One of the science departments stopped ordering articles via EDD because the quality of the fax was so poor, preferring the traditionally slower but good quality photocopies provided by the British Library. On contacting Blackwells about this, they explained that clarity of small print, footnotes, diagrams and graphs always suffered with current fax technology.
- The service had been advertised as a 24 hour service and delays of up to 48 hours became quite frequent. Other minor irritations occurred such as articles arriving with pages missing. These were immediately re-sent once we contacted the Help Desk. We even received some articles that were not requested by us. These, I am pleased to say, were not charged for.
- From a Library point of view, our major problem was educating the users that the trial was for articles required urgently and not an electronic substitute for Inter Library Loans (ILLs). Many users tried to take advantage of the free service rather than using their School's ILLs budget whilst others preferred to send and receive their requests electronically, rather than coming over to the Library and filling out an ILL form. These problems were reduced considerably after a letter was sent out to the departments reinforcing the aims of the experiment, and on a couple of occasions, some requesters were contacted in person.
The Way Forward
The EDD steering group noted that:
- to a large extent the experiment had demonstrated that EDD could be a viable alternative to expensive periodical subscriptions but further research was necessary before Departments would be willing to risk cancelling journals and relying solely on EDD and ILLs.
- an EDD service should be provided in the future as a complementary service to ILLs and not a substitute for it. Although immediate delivery was not always necessary there was a demand for it, and users would be unhappy for the service to be discontinued.
- research into other EDD service providers on the market should be carried out. Investigations should concentrate on cheaper alternatives and/or better quality service.
- administratively, the service was too time consuming to handle on a busy enquiries desk. Any future trials should be handled by the User Services team. This however would have long term implications on staffing and time. Other options to be looked into could involve the departments ordering their own material but again would lead to monetary handling implications.
- developments in technology were advancing so fast, that with the increasing number of full-text journals becoming available online, would this type of service be needed in the future?
Alternative document delivery systems : a survey
As a result of the above recommendations, an investigation into other document delivery services on the market was undertaken in Spring 1996.
The following criteria were used to determine which services best suited the needs of the user, and those of the Library staff carrying out the trial.
What do our users want?
- 24 - 48 hour delivery
- direct personal delivery i.e. via email, fax, post, without a visit to the Library
- good quality, clear copies of articles requested
- to request articles either by email, post or in person
What does the Library want?
- an efficient, cost effective, fast service
- to be able to request items via email, fax or phone, without having to search a database first
- to receive "confirmation of receipt" of the request
- good back-up service/customer support
- easy payment methods
Research into the other service providers on the market, such as BIDS, Firstsearch and EBSCOdoc, showed that they worked in a similar way to Uncover; they were all rejected because the initial searching had to be undertaken by the Library which was a time consuming and lengthy process.
The following services were recommended for further trial. In all three cases, the service providers acted as the intermediary and carried out the search for the requester upon receipt of the request:-
- the British Library (BL) Urgent Action Service
- Delft University of Technology Document Delivery Service (TU DELFT) and
- the Royal Society of Chemistry (RSC) Photocopying Service
After discussions with the User Services Librarian, it was felt that:-
- the RSC service was very subject specific and should not be used and that the British Library Urgent Action Service and TU DELFT should run in tandem. (An outline of the three services are given below).
- as the already very busy ILLs staff would be running the new trial, it would be better to use one familiar service - BL Urgent Action, and one totally new service - TU DELFT.
- results of the previous trials showed that the majority of requests had been made by science departments, and that TU DELFT would be able to accommodate most of their requests. Any requests unfulfilled by TU DELFT would then automatically be sent to the BL whose coverage is far more comprehensive, but whose cost is more than twice the price of TU DELFT.
- this latest trial, using both services, could start in August 1996 and continue until funds ran out. (A sum of 2,000 pounds had been allocated to run this trial).
- No. of titles: Approximately 50,000 per year
- Coverage: Multidisciplinary
- Speed of delivery:
- Fax - same day delivery
- 1st class mail - within 24 hours
- Requests received before 3.30 guarantee same day despatch
- Delivery via:
- Articles can be sent to any address given and still be charged to the Library account.
- Ordering via:
- Electronically using ARTTEL 2
- 1st class post (article up to 50 pages) = 3 BL vouchers @ 4.50 pounds = 13.50 pounds
- Fax = 4 BL vouchers @ 4.50 pounds = 18.00 pounds
- Payment: Would continue to use the existing method of BL vouchers. There is no longer a charge for unsuccessful requests.
- Quality of copies received:
- Photocopies sent 1st class should be of the same good quality as the copies sent out by the BL normal service
- Fax - dependent on quality of fax machine. Not good for diagrams and/or graphs
- Customer support: Will update customer on findings within 2 hours of receiving request.
- No. of titles: 13,000
- Coverage: Scientific and Technical literature.
- Speed of delivery: Rush services (request dealt with that day usually within 8 hours of receipt)
- Delivery via:
- Ordering via:
- Price: Cost per article sent by ftp = 11.20 guilder/gilder + VAT = ú 5.19
- Payment: Setting up a deposit account was preferred, so the costs of the article could be deducted
- Quality of copies received:
- Fax - depends on quality of fax machine. Not usually good for diagrams and/or graphs
- ftp - printed off Library laser printers - good for graphs and diagrams
- Customer support: Emails are sent confirming they are processing the order. Cranfield University had used the service and were very pleased with it.
- Internet address: http://www.library.tudelft.nl/
- No. of titles: 2,000 (700 current)
- Coverage: Chemistry and related areas
- Speed of delivery: Express service - items can be faxed on the same day if order received before 4.00p.m. or despatched by first class post on day request received.
- Delivery via:
- Post urgent action - 1st class same day dispatch
- Fax urgent action - faxed same day as receipt
- Ordering via:
- Fax (by voucher or typed request)
- Urgent action post = 2 vouchers = 7.40 pounds
- Urgent fax = 4 vouchers = 14.80 pounds
- Payment: Annual corporate membership: 145 pounds. Members buy a supply of pre-paid vouchers in books of 10 and use as requisition forms.
- Quality of copies received:
- Fax - dependent on quality of fax machine
- Post - depends on quality of photocopying - good for diagrams and graphs
- Customer support: Cranfield University also used this service and were pleased with it.
- Holdings: A full list is available on request
- Internet address: http://chemistry.rsc.org/rsc/library.htm
Baker, David. Document delivery: the UEA experience. Vine 95, 1995, pp. 12 - 14
Beddall, Jane. Carl Uncover service. Managing Information, 1(10), 1994, p. 50
Dade, Penny. Pilot trial of Blackwell's UnCover database. Vine, 98, 1995, pp. 40 - 42
Dade, Penny. Electronic information and document delivery: final report on the pilot trial of the Uncover database. Vine, 103, 1997, pp. 43 - 45
Thanks to Hannah South, Subject Librarian at the University of Bath, for her help in reviewing the practices of other service providers on the market.
University of Bath
Phone: 01225 826826 ext.5248