Electronic Document Delivery: A Trial in an Academic Library

Following the original Electronic Document Delivery (EDD) trial using Blackwell’s Uncover Service, a number of recommendations were made which led to an investigation into other document delivery services available on the market. The criteria used to decide which services would best suit the needs of the user, and the University of Bath Library staff who would be carrying out the trial are outlined in the first article of this two-part series (Williams).

Two services were recommended for further trial, the BL Urgent Action Service (BLUA) and the Delft University of Technology Document Delivery Service (TUDELFT). BLUA was used as it was already familiar to the library staff and because of its extensive coverage. TUDELFT was decided on because of the low cost and because the majority of requests in the past had been science and technology requests. It was hoped that TUDELFT would be able to accommodate most of the requests.

Aim

• to run the two services in tandem to compare and contrast the cost, speed of delivery, and the quality of copies received
• to compare this trial with the Uncover trial and decided which service/s best suited the needs of our users and the Library.
• to look at the advantages and disadvantages of all the services
• to use the trial to see if EDD could save money on expensive journal subscriptions
• to continue to quickly obtain “urgently required documents” - a service which had become very popular
• to provide a complementary service to the ILL service and not a substitute for it

Funding

A total of £2,000 was made available from Library funds to run the service initially for six months starting August 1996. If the service became so popular that it became administratively difficult to run, the service would be withdrawn immediately. The trial eventually finished at the end of the Easter vacation in April 1997.

Method

• The trial was open to all staff and postgraduates within the University. It was emphasised as a service which was available for urgently required articles only, and not as a substitute for normal ILLs. The trial was advertised in staff and student newsletters, as well as the combined Computing and Library newsletter. Fliers were also put in staff and postgraduate pigeon holes within the Departments.
• The trial was run by one member of staff from the User Services team. Applications were accepted by email, post or in person.
• Approximately 90% of the requests were sent to TUDELFT first, unless they were very obviously Arts & Humanities based. If TUDELFT were unable to process the order, the request was immediately sent to the BL for Urgent Action.
• Requests to TUDELFT were sent by email and received via ftp (file transfer protocol). The articles were then printed out on the library laser printers. Requests to the BL were faxed and articles were sent out by first class post.
• Requesters were informed by email that their requests had arrived which they collected from the Library; the copyright declaration forms were signed on collection.

Review after 4 months

A review of the service took place in December 1996. Overall it was felt that the service was running well.

In the 4 months since the trial began:-
137 requests had been processed; 7 had not been supplied and 6 were returned as they were already on our shelves; 10 had not been collected; 58 staff/postgrads used the service; 115 titles had been provided; 50% of the £2,000 allocation was spent

Feedback

A questionnaire was sent to all users in December 1996 - ¹⁹⁄₄₅ replies were returned.

Results of the whole trial

1. Speed of delivery

The following table illustrates the length of time taken to process and deliver requests by the 2 service providers. In total 271 requests were made. 74.6% of those provided by TUDELFT and 56.4% provided by the BL were delivered within 48 hours.

However, of the 271 requests made (the majority of which were science and technology requests) TUDELFT were only able to provide just over 50%. At first they were able to provide a very high percentage of the requests, but as they trial continued this reduced to virtually a ⁵⁰⁄₅₀ split.

Types of journal titles supplied

Details of borrowers by department and numbers of requests

The statistics showed that in the 9 months the trial ran, 271 requests were made by 85 different people. The two highest requesting Departments were Biology and Chemistry; most of these requests were satisfied by TUDELFT. The third highest requesting school was Management; most of these requests were satisfied by BL Urgent Action.

Considering the high proportion of science and technology articles requested, it was disappointing that TUDELFT only managed to provide just over 50%.

Advantages and disadvantages of BLUA, TUDELFT and Uncover

1. Cost effective

At £5.19 per article TUDELFT were 2½ times cheaper than the BL and about 3 times cheaper than Blackwells. BL’s and TUDELFT prices are guaranteed for articles under 50 pages, whereas Uncover’s prices are dependent on the individual copyright charges made per article .

2. Delivery via email, fax or post

BL articles were sent out by first class post to the Library, so we were dependant on the postal service - but overall found the service very effective. TUDELFT articles were sent via ftp which were received by the Customer Services Librarian and then sent to the Library laser printers. Uncover was the most popular service with the users because the articles were faxed directly to the requester as required.

3. Quality of articles received

Very good quality from both BL and TUDELFT. One of the main criticisms of Uncover had been the quality of the faxes particularly for articles with diagrams and tables.

4. Request articles via email, fax or phone

Most requests were made by email which were printed and then faxed to BL, or forwarded straight to TUDELFT. Uncover had its own request system, which was very time consuming as all the information had to be re-keyed.

5. Database searching

All requests were directly sent to BL or TUDELFT without having to search a database first to check if the order could be processed. With Uncover, the database had to be searched first, which was one of the major problems with the service, because searching and access was so slow, and very frustrating.

6. “Confirmation of receipt” of requests

Uncover did not provide this service at the time the trial was being carried out. BL confirmed whether or not they could process the request within a couple of hours of receiving the order. TUDELFT were variable, sometimes confirming quickly, other times not actually confirming the order was going to be processed until after the article had arrived. This caused some problems as we were sending most of the orders to TUDELFT first and then onto the BL if they could not process the order. This was particularly frustrating for the users who required the articles urgently, .

7. Ease of payment

BL payments made by the voucher scheme so no increased workload. Accounts were set up with TUDELFT and Uncover who debited them after each transaction.

8. Copyright declaration

Copyright declarations were required for all requests from the BL and TUDELFT. The forms were signed by the requester on collection of the article from the Library. Uncover were able to fax articles directly to the requester because a copyright charge per article was included as part of the service.

9. Customer support

All 3 services have a very good back up service, although as the trial continued, the responses from TUDELFT were very variable. In some cases the article arrived before the message confirming that they were processing the order. Near to the end of the trial, it was felt that TUDELFT was getting slower in replying and providing articles.

10. Administrative handling

The BL service has been used by the University of Bath for a number of years, and therefore we are used to the administration required.

TUDELFT was administratively more difficult to set up initially because all the procedures were done electronically. This took time to get used to and a few procedures needed to be “tweaked”. It was also time consuming getting the article because of the number of commands needed to retrieve the articles from the server. A macro however could solve this.

Both the BL and Uncover services sent the article in a format which could be used immediately.

TUDELFT became a very paper administrative service. All the information about a request needed to be on paper to help with the customer service side of the service.

Uncover was frustrating because of having to rekey the information and search the database, but overall was very easy to handle administratively.

Technical problems

With TUDELFT we had some problems which were peculiar to OUR environment.

1. Articles were being sent by TUDELFT via ftp. To print these tiff format files, they had to be converted to postscript using the tiff to postscript command. Large postscript files were created but were very slow to print (approximately 10 minutes per page) because of the serial line . To rectify this, the files were converted to postscript in a two stage process, creating smaller postscript files which improved the download time.
2. Each page was received individually via ftp. These were sent to the Library laser printers individually which as already mentioned was extremely slow. Computing Services wrote a program to assemble all the pages into one document, which was then sent to just one printer. It still took a long time for the pages to be printed but it only kept one of the three library printers out of action.

The Way Forward

The feedback from the trial showed that the main uses of EDD are:-

• to quickly obtain urgently required documents
• an alternative to the ILL service by some users
• an option when cancelling expensive journal subscriptions if necessary (one of the Library’s main aims when it first started the EDD trials)

The results of the trials suggest that BLUA is an efficient and easy to use service, but expensive. Although initially difficult to get used to, 50% of the requests were processed by TUDELFT, with 75% of those requests being processed within 48 hours. In balance the comparatively low cost of the service outweighs the potential problems. Although Uncover was popular with the users, access and searching the database was too slow, it is expensive and the poor quality of the faxes make it an unacceptable service at the present time.

The University of Bath Library staff would like to develop an EDD service further, but the costs of implementing and integrating new procedures into the ILLs department is not feasible at present with the possibility of a new library system being implemented in 1998.

However once implementation of a new system takes place, it is recommended that EDD is offered as part of the ILLs service.

Until then, starting in September 1997 the following service will be offered to the Departments:-

1. Urgent requests will be accepted by the Customer Services Librarian, via the Web page or on a signed ILL form with URGENT ACTION written on the front.
2. A copyright template will be available on the Web which must be downloaded by the requester, completed and signed and put in the internal post to the Customer Services Librarian.
3. The Customer Services Librarian will immediately request the document from TUDELFT or BL Urgent Action, depending on whom she thinks will be able to supply the document. If TUDELFT are unable to supply, the request will then automatically be sent to BL.
4. On receipt of both the article and the signed photocopy declaration, the article will be sent by internal post to the requester.

Implications for the Departments:

• the EDD trials are over, the service will no longer be free. The costs will be charged to the Departments and they must decide how they want to use the service.
• costs are as follows:
  BL Urgent Action service is the equivalent of 3 ILL forms (£13.50); TUDELFT is £6.50 (i.e. cost of the service plus printing charges).
• methods of payment:
  BL Urgent Action requests will be charged as normal; money will be credited to the TUDELFT account which they will draw upon once the requests are fulfilled. Once a term the costs will be worked out for each department and be adjusted on the Library spreadsheet.

References

WILLIAMS, F (1997) Electronic Document Delivery - a trial in an academic library.
ARIADNE issue 10, July 15
[WWW] http://www.ariadne.ac.uk/issue10/edd/

Acknowledgements:

My grateful thanks to Pat Heynes, Customer Services Librarian, University of Bath for all her hard work in running the trial using TUDELFT and BLUA and to Dennis Davis, BUCS for his technical help.

Author details