UKSG Serials Automation Survey 1989

The third UK Serials Group Automation survey was conducted in December 1988 and January 1989 and the detailed, four page questionnaire was sent to a total of 254 libraries. All UK library members of UKSG were approached; in addition, a number of libraries who sent delegates to the joint UKSGINational Acquisitions GroupIMARC Users Group exhibition of automated library systems in" Birmingham in November 1988 were also approached. The reason for including these additional libraries was to broaden the scope of the survey, particularly to public libraries, which are under represented in UKSG membership. Replies were received from 123 libraries giving a response rate of 48%. The breakdown by type of library was 70 academic libraries, 18 government libraries, 15 public libraries and a small number of medical, research and industrial libraries. The largest proportion of respondents (57%) were academic libraries. Due to the low response rate in some categories it was clearly difficult to identify and predict trends.


Introduction
The third UK Serials Group Automation survey was conducted in December 1988 and January 1989 and the detailed, four page questionnaire was sent to a total of 254 libraries. All UK library members of UKSG were approached; in addition, a number of libraries who sent delegates to the joint UKSGINational Acquisitions GroupIMARC Users Group exhibition of automated library systems in" Birmingham in November 1988 were also approached. The reason for including these additional libraries was to broaden the scope of the survey, particularly to public libraries, which are under represented in UKSG membership. Replies were received from 123 libraries giving a response rate of 48%. The breakdown by type of library was 70 academic libraries, 18 government libraries, 15 public libraries and a small number of medical, research and industrial libraries. The largest proportion of respondents (57%) were academic libraries. Due to the low response rate in some categories it was clearly difficult to identify and predict trends.
Respondents collection size and expenditure ranged dramatically from large national collections through to small special libraries, thus giving an interesting overview of different types of serials automation. 42 of the 123 libraries which replied had no serials automation at all, but many were actively evaluating systems. Of the 81 libraries (65%) which had some form of serials automation, automated processes ranged from cataloguing only, to fully automated systems handling check-in, claiming, management information and OPAC. Figure 1, shows the extent of serials automation in all responding libraries.
A wide variety of systems were represented, including fully integrated commercial systems, subscription agent's packages and in-house development. Figure 2, shows the various types of serials automation implemented by responding libraries.
In-house systems still dominated the serials scene although it is likely that this situation will change significantly over the next few years.
During analysis of the data a number of interesting issues arose from respondents' answers. There appears to be some inconsistency in the "type of library" category. For example, some medical libraries identified themselves specifically as medical libraries, others selected the academic library option. Some confusion also arose between the government and national library categories. The definition of a serial was raised by question three which asked for the number of current serials taken. Certain answers in this section were verified by telephone calls and it became apparent that while some libraries take the broadest possible definition of a serialincluding such items as official government publications, technical reports, statistical series etc. within their collections, others separate out special collections. Clearly therefore, when comparing size of collections we are not always comparing like with like. On the subject of expenditure, the questionnaire identified six blocks of expenditure, ranging from up to f30,OOO to f200,OOO and above. A significant number of respondents selected the last box which, on reflection, was pitched too low. Another question which brought a surprising response was that relating to electronic mail. Only four libraries in total admitted to using electronic mail to contact subscription agents for orders, claims, messages etc. Examples quoted included Data Swets, Dialcom and Fax. We might have expected this figure to be higher particularly with the rise in popularity of the fax machine.

Academic Libraries
Academic libraries represented the largest proportion of respondentssome 57%comprising universities, polytechnics and colleges. Replies were received from 46 universities (including Colleges of London University), 12 polytechnics and 12 colleges; a total of 70 completed questionnaires.
As expected, university libraries had the largest collections of serials and the highest number of current titles. No distinction was made between current titles and current subscriptions. 12 university libraries received more than 5,000 titles. Figure 1.

Automated processes
Copyright libraries appeared near the top of the list but the library which received the highest number of current titles (13,000) was not a copyright library. 4 university libraries took between 4,000 and 5,000 titles and 7 between 3,000 and 4,000. All polytechnic libraries which responded to the survey (12) took between 1,000 and 3,000 titles as did 13 university libraries. The majority (66%) of college libraries received less than 1,000 current titles. There appeared to be no correlation between single and multi-site libraries and the number of serials taken. 34% of respondents operated multi-site libraries.
seemed to be in the forefront, with only 2 having no cataloguing system. Question ten asked "In what format are the serials records held (eg. UK MARC)?" This produced a variety of answers such as in-house, AACR 11, Non MARC as well as UK MARC. Whilst indicating that automated systems were used in their libraries, a number of respondents failed to complete this question. 38% of libraries used UK MARC format; a relatively small number when one considers the increasing pressure to co-operateboth locally and nationally in the area of serials acquisition and de-acquisition. Certainly a good proportion of all but one were university libraries. Ten libraries spent between fl50,000 and £200,000. Only 8 academic libraries spent less than £30,000.
The amount of money spent on serial subscriptions reflected a similar distribution to the number of current titles received. 33 academic libraries out of a total of 70 spent more than £250,000 on serial subscriptionsa massive 47%. Of these 33 libraries Academic libraries displayed a fairly high level of automation in general; 60 of the 70 respondents had some form of serials automation or were actively planning to introduce a system. Not surprisingly, serials cataloguing was the most frequently automated activity, with two thirds of libraries having either a batch or online cataloguing system (online predominating). Clearly without an automated cataloguing system other automated projects are not possible. Polytechnic libraries academic libraries (43%) were participating in union lists; it should however be pointed out that 29% were members of library co-operatives, thus automatically entering records into an established co-operative database.
The rise in popularity of the online public access catalogue (OPAC) was evident. Exactly 50% of university libraries and one third of polytechnic and college libraries indicated that OPAC was available within their institution. It is probable however that only serial bibliographic records are available on most library OPACs as the integration of serials check-in data is still under development in a number of major commercial systems. The number of libraries using an automated check-in facility was still relatively low. Only 8 libraries had online check-in: 5 universities, 1 polytechnic and 2 colleges. Of these 8, even fewer used the automated explained by the fact that very few academic libraries actually route serials. Routing was found to feature more significantly in the special library categories. The survey might also have been expected to show lending as a marginal activity as academic libraries do not traditionally loan serials. However, 19 libraries (80% of which were universities) had automated serials lending, and a majority indicated that their system was not an integrated library system. It must therefore be assumed that lending was a function of the serials automation package. system for claiming (6) or routing (4). The small number using the routing facility might be Binding seems to be one of the last processes to be automated. 4 universities and 1 college had binding records in machine readable format. This might be explained by a number of factors. First and foremost, many commercial systems have not yet developed a binding module for libraries to use. Binding automation is also a laborious and staff intensive task which cannot be implemented until the serials bibliographic database and check-in records are in place. Given these circumstances, it is not surprising to find that binding currently receives a low priority in the hierarchy of serials number of libraries are using fax as a speedy and accurate method of communication.
issues and back issues was not practised in many" libraries. This may be due to the fact that libraries are not placing large numbers of new orders in the present economic climate and it is as easy to handle that process manually. Only 9 libraries had automated ordering. Subscription agents might be interested to note that electronic mail did not feature significantly in the day to day running of the serials department. Only 1 academic library was using a subscription agent's electronic mail facility. Similarly, only 1 respondent mentioned facsimile transmission although it is likely that a larger Online ordering of new serials titles, replacement Management information and financial control are key areas in any automated or manual library system; in serials management they are of paramount importance due to the complicated nature of serial acquisition, invoicing and payment. Moreover, academic libraries are under increasing pressure to justify their existence and it is in the areas of serials acquisition and collection development that attention is being focused. One of the major advantages of an automated system is that data can be manipulated in a variety of ways to ' produce management reports. The survey showed I automation. Figure 2. Vol. 3, N e 1, March 1990 UK9G Survey that automated financial control took priority over general management information. 34% of libraries had automated financial control compared to 23% who had automated management information. Of the 24 libraries which had automated financial control, 17 had online facilities and 7 batch. Similarly in the area of management information more libraries had online access as opposed to batch processing. This does however raise the question of how management reports are generated within the library. It is, of course, possible to have online access and online updating facilities on a system but to only run reports on a weekly or monthly basis. Such reports may well be considered "batch".

Serials -
27% of responding academic libraries, had no automated serials system. Just over 28% of libraries had in-house developed systems (surprisingly the highest group). Co-operative systems were represented by BLCMP, SWALCAPILibertas and OCLC, and accounted for almost 29% of systems. As expected, subscription agents' systems did not feature significantly within academic libraries. Most agents' systems do not handle the large numbers of titles present in academic library databases. Other commercial systems in operation in academic libraries were ADLIB, Dynix, GEAC, IBM/DOBIS/LIBIS and Sydney Microlibrary. Of the 51 libraries with an automated serials system, 22 of those systems formed part of an integrated library system. This is clearly the trend within large academic libraries. This is reinforced when we turn to the question of future plans. A high proportion of libraries with no automated system or in-house developed systems are actively considering purchase of an integrated library system. Another significant development, frequently remarked upon, will be the full implementation of the serials modules of co-operatives such as LIBERTAS and BLCMP. Not one academic library claimed to be evaluating an agent's system and only a very small number intended to continue in-house development.

Government/State Libraries
Government and State libraries accounted for some 16% of the respondents, a total of 20 libraries in all. The size of the libraries' current serials collections varied widely, as expected. 5 libraries received less than 1,000 titles, with a further 9 subscribing to between 1,000 and 4,000 titles. The remaining 6 ranged from 10,000 and 94,000 titles each. There was no apparent correlation between the number of titles taken and whether the library was single or multi-site. This random pattern was not reflected in the amount of money spent on subscriptions. Tho main groups were identified, with 30% of libraries spending between f30,OOO and f75,000, whilst 45% spent in excess of f200,OOO.
75% of libraries in this sector had some form of automation. Of these libraries, 9 had automated cataloguing (online predominating) and check-in, and 10 had some form of automated financial control. As would be expected routing and management information were of considerable importance, with 9 libraries automating these activities. Binding, lending and OPAC applications were least represented, with approximately half of the sample having automated claiming and union lists. The most significant general trend was the very high degree of online applications, with very few applications being run in batch.
Of the 14 libraries who had automated some serials applications, only 3 had integrated library systems, the majority of libraries opted for micro-based systems. Of these 14, 5 had implemented systems developed by agents, another 5 had systems developed by commercial suppliers, the remaining 3 had developed in-house systems. Only 4 libraries in this category employed a consultant to assist with the implementation of automation, 3 libraries opting for external consultancy. Surprisingly none of the libraries currently use electronic mail, and only 1 library has a link to an agent's database. 3 libraries planned to install a system during 1989, whilst another 6 were considering installing systems in the short to medium term. Only 1 of these had plans to install an integrated library system.

Public Libraries
Public libraries made up 12% of the libraries responding. Most serials collections were small in 'comparison to other sectors of the library community; 7 libraries subscribed to less than 1,000 titles, with a further 5 subscribing to between 1,000 and 2,000 titles, only 2 libraries subscribed to more than 5,000 titles, the largest figure being 12,000. This pattern was mirrored by the expenditure on current subscriptions. 7 libraries spent less than f30,000 pa., another 4 spent between f30,000 and f75,000. The remaining 4 libraries' expenditure was spread across the range, with only 1 in excess of f 200,000.
With serials collections generally being relatively small in public libraries, the incidence of automated serials applications might be expected to be comparatively low. Indeed, from the sample of 15 respondents, only 2 reported any automated applications. Of these 2, 1 had only automated cataloguing, the other having a combination of batch and online applications for cataloguing, ordering, union listing, routing and lending. Both systems were developed in-house and neither were part of an integrated library system. Consultants were not employed at any stage and none of the libraries in the sample used electronic mail, or had any links through to an agent's database.
1 library expressed plans to purchase a system from a subscription agent and another 3 had plans to introduce serials control as part of an integrated library system. 2 libraries were currently evaluating systems.

Industrial Libraries
Although industrial libraries are well represented in UKSG membership there was a disappointing response rate in this category. Of the 5 libraries which replied (4% of the total respondents) the number of titles subscribed to in each case was less than 1,000. The amount spent on subscriptions however varied greatly, with 2 libraries spending up to f75,000, another 2 spending between f75,000 and f100,000, and the remaining library spending between f100,000 and fl50,000 annually. It was difficult to determine the reason for this wide fluctuation. Only 1 respondent reported having more than one site, so no relationship between sue of organization and expenditure could be inferred.
Only 1 library had yet to introduce automation of any serials applications. Of the remaining 4, 3 had automated OPAC, ordering and financial control, 2 had automated cataloguing, check-in, claiming, binding, union listing, routing and management information. Only 1 library had automated lending of serials. Of the 4 automated libraries 2 had integrated library systems; 3 had introduced systems developed by commercial suppliers, the fourth had developed an in-house system. None of the libraries reported using electronic mail, but 1 did indicate that they had a link to an agent's database. 1 library planned to install an agent's system during 1989, while another 2 were actively evaluating systems.

Medical Libraries
Of the 5 libraries replying in this category 4 subscribed to less than 1,000 titles, the remaining library taking between 2,000 and 3,000 titles. 2 libraries spent less than f30,000 annually on serials, another 2 spending between f30,000 and f75,000, with the remaining library spending between f 150,000 and f200,000.
Serials automation clearly had a low profile in this particular sample. Only 1 library had automated any of its serials activities, running an integrated library system supplied by a commercial supplier. Electronic mail facilities were not used and none of the libraries had any link to an agent's database. 2 libraries were considering installing integrated library systems in the short to medium term, whilst 1 library stressed that serials applications were not a priority.

Research Libraries
Research libraries accounted for nearly 6% of the respondents, some 7 libraries in all. Numbers of current titles subscribed to varied widely, with 5 subscribing to less than 1,000 titles, another taking between 1,000 and 2,000 titles, and the seventh subscribing to 10,000 titles. Expenditure on current titles was spread evenly across the lower ranges. 3 libraries spent less than f30,000 annually, another 3 spent between f30,000 and f75,000, and 1 library between f75,000 and f 100,000.
In an area where a high incidence of automation could be expected, 2 libraries had no automated serials applications. Of the remaining 5, 4 had automated cataloguing. 3 libraries provided management information capabilities and 2 provided check-in, claiming and union lists. OPAC, ordering, financial control, routing and lending were only automated in 1 library each. Of the 5 libraries who reported having automated some serials applications, 3 had integrated library systems. 3 of the 5 systems were developed in-house, 2 were purchased from commercial suppliers and the fifth was an agent's system. 2 libraries employed a consultant to assist with the implementation, and in both cases an internal consultant was used. Research libraries utilized electronic mail to a greater extent than any other category (3 out of 7), and 2 also had a link to an agent's database. 1 library planned to install an agent's system during 1989, whilst another 2 were in the process of evaluating systems.

The database
Any member who wishes to see the complete database, or sections of the database sorted by type of library, are asked to contact Hazel Woodward on Loughborough (0509) 222352. It should be noted, however, that in certain entries the name of the responding library will remain confidential. O