The purpose of this paper is to show that there are small differences in the methods used in a public or special library, when they want to carry out a performance measurement. To me it seems as almost every indicator is applicable in all types of libraries.
On the other hand the interpretation of the indicators will differ because the stock and user groups are different. As the choice of indicators has to be linked to the goals, mission and objectives o f the library, the choice will vary from library to library. This link will also have an influence on how the library will use the collected data and indicators in the evaluation and future planning process.
Another intention of this paper is to try to reduce the boundaries between the two sectors inside IFLA. In my opinion it will be of value to let the discussion on performance measurement, cover all l ibrary types. Last year, at the Cuba Congress, a discussion group on performance measurement in academic libraries was established. This is a very positive initiative. In my opinion this group should not be limited to academic libraries, but cover all library types. I would like to say: Libraries are different, but except for the sector and user group, there are also differences inside the same type of libraries, e.g. size of the library, and whether the collect ion is on open or closed shelves.
While traditional statistics focus on the library itself, performance measurement will through its focus on output and outcome have the user in the centre. In performance measurement we try to clarif y the link between input, output and outcome in order to make a systematic improvement of the library services. It is important to stress the user view because this point establishes many of the simi larities. When we change focus to internal operations and organisation the differences become larger.
The user group of a public library is heterogenic, and in a special library more or less homogenic. This will influence the goals and objectives of the library, but not the choice of methodology. To me it seems as the difference between library types is greater in traditional library statistics.
Performance measurement is not a clear concept. There is still a lack of clear definitions. To me qualitative measures fall outside, even if this can tell something about the performance. I mean the same about pure user surveys. If you link the data of the user survey to other indicators to get an overall picture, then it is a part of performance measurement.
Performance indicators are created by comparing quantitative data elements in different combinations. The purpose of the indicators is to analyse data in order to clarify the output and outcome of th e library services and see how well the library is performing.
The data elements have to be chosen from internal systems, by manual counting or observations, external sources or direct form users. These elements are more or less present inside all kind of librar ies.
What is different - and how will it influence the measurement The services offered by a special library are more specialised, and they will often need to study each part of their service more in detail. In general, performance measurement will be a more complex task in a special library if they want to evaluate the whole service.
The user group is different, but for both library types it is of interest to the percentage of registered users as a potential of potential users, and to see if there are some over- or under represe ntation of certain groups. A more educated and demanding user group makes the need for more indicators to track the services offered.
The use of the collection and the collection itself are different. Loan is more important to the public library, as in-house use and photocopying are more common in special libraries. Every library s hould track both to get an overall picture. E.g. turnoverrate and usage by category are important measures to all libraries. The fact that some special libraries have other missions than serving the user, as deposit and preservation tasks, make it more difficult to analyse the data.
Special services tailored to individual users, as SDI, are more common in special libraries and have to be analysed on their own. In general they have a possibility to choose which services they want to offer different user groups.
Use of technology and use of electronic sources and delivery are more common in special libraries, but do not change the overall picture. Automated library systems have the possibility of generating lots of data about the use and collection, but it is a fact that many system developers have not paid to much attention to this area.
The library network plays a more important role in special libraries, especially inter library lending and borrowing, but this is not a service that is unique to this kind of library. Some libraries may have tasks and collections that have the whole library network as the target user group.
Finally a special library often will have more skilled personnel with a theoretical background on statistics and user surveys and other methodology inside their own institution to do this kind of wor k.
The staff has to go through the same planning process before they are starting a performance measurement. The process on collecting the data is almost the same.
Library users are executing the same operations inside the library, even it is likely that public library users have a higher level of selfservice. This may turn the focus on indicators of the use of the collection.
As a conclusion I would say that it seems easier to make central guidelines and share general experiences on performance measurement in public libraries, but these are of clear value to all libraries . If this discussion on performance measurement is not going between the two library worlds, both sectors will lose out. There has been a tendency for this, e.g. that quite similar discussion has taken place inside the sections of academic and public libraries.
To me it seems that the public libraries are stressing the user focus and want to get an overall picture of their service, more than the special libraries. On the other hand, because of the complexit y of a special library's services and collections, special libraries often make a more detailed and sophisticated study of some parts of the library service. These experiences should be shared with t he library community as whole.