|Bridge Builders: African Experiences with Information & Communication (BOSTID, 1996, 304 p.)|
|Case studies on the collection, management, and dissemination of local information resources|
by Stella Monageng
Stella Monageng is the Senior Documentalist at the National Institute of Development Research and Documentation at the University of Botswana. She has wide experience in the development and management of computerized databases and has been working in this area of expertise since 1983.
BACKGROUND AND CONTEXT OF THE PROJECT
This case study describes my experiences in developing databases at the National Institute of Development Research and Documentation (NIR) of the University of Botswana. I will discuss a project entitled Development Sciences Information System for Botswana (DEVSIS-Botswana), which is the system we created to facilitate the collection, organization, and dissemination of Botswana's socioeconomic development information.
DEVSIS is a project that dates back to the mid-1970s, when a number of international agencies, including the International Development Research Centre (IDRC) of Canada, established a system that would help meet the information needs of planners and decision-makers responsible for economic and social development planning. The concept was never implemented as the global scheme that had been envisaged. Instead, a number of national and regional development information systems evolved out of the program. DEVSIS-Botswana is one example of such a national information system. In collaboration with the Pan African Development Information System (PADIS), the NIR launched our project in May 1984 with the financial assistance of IDRC.
The National Institute of Development Research and Documentation (NIR)
The NIR was established in 1975 as a documentation center. In 1978, it added a research component. Its objectives are to:
a) Coordinate and conduct research on issues of socio-economic
and environmental development affecting Botswana,
b) Develop the national research capacity within Botswana; and
c) Publish and disseminate the results of the research conducted.
The research activities focus on agriculture and rural development; education; environment; health and nutrition; Basarwa1; and women and gender issues.
The Documentation Unit of the NIR implemented the project. The Unit was established in 1975 and later developed into the present research institute. Its main objective is to collect, process, and disseminate unpublished or grey literature on or about Botswana. The Documentation Unit has a library through which the information that is collected and processed is disseminated.
The Documentation Unit and its library deliver the following products, technologies, and services:
· Reference Service through which we assist users in accessing the library's computerized databases. · Selective Dissemination of Information through which we offer a current awareness service to the Institute's researchers through their research interest profiles. · Newspaper Clipping Service through which we clip and file key articles falling within the research areas of the Institute. · Referral Service that links NIR with other major information centers such as the National Library National Reference Collection, the University of Botswana Library Botswana Collection, and several Government department libraries. · Photocopying Service, which is provided for a fee. This service is not only used locally but is also used by a lot of researchers from outside Botswana. · Computerized Databases maintained by the library, the first of which was set Up in 1986. Until quite recently, when the University of Botswana Library computerized its operations, the NIR Library was the only one with a computerized information system. This attracted a lot of researchers because it made retrieval time shorter and helped to inform the users whether or not certain information was available. · Devindex-Botswana. which is an output of the computerized databases. Documents are collected, cataloged, abstracted' and indexed and then in put into the computer. When a reasonable number of these have been processed, an index is produced.· Technologies used include: electronic mail, through HealthNet and Internet; Facsimile; and CD-ROM. The Documentation Unit has been selected as a POPLINE site for Botswana and will therefore have the POPLINE CD-ROM database.
History of the Project
In 1981/1982 many information professionals were trying to establish an information system called the Southern African Documentation and Information System (SADIS). The envisaged aim of the system was to assist countries in the region to build up their information infrastructures as a basis for a coordinated regional information system. Unfortunately, the system never got off the ground.
Botswana, however, through the initiative of the NIR, decided to set up its own socioeconomic development information system. We approached PADIS for assistance in computerizing our documentation system and also in acquiring funding for the exercise. The NIR circulated a proposal among various government ministries for comments and suggestions. The responses received were so positive that the Ministry of Finance and Development Planning approved the proposal and agreed to submit it to IDRC for funding.
NIR decided to maintain the momentum started by the SADIS efforts for several reasons. First, we realized that the situation regarding the documentation of Botswana's socioeconomic development was far from satisfactory; much development information was generated but it was ending up in very inaccessible places. We also found it necessary to establish such a system so that Botswana could benefit from development information available in other countries. (See Box 1.)
Objectives of the Project
The principal objectives of the project were to organize the national economic and social development information in support of the planning process in Botswana and to strengthen Botswana's capacity to participate in the regional PADIS network and in the proposed SADIS program. The specific objectives of the project were to:
a) Collect, organize and disseminate Botswana's national
information and documentation related to its economic and social
b) Participate in the regional PADIS network;
c) Train staff in the organization and operation of a specialized documentation center; and
d) Determine the computer equipment, staff and training that would be needed in order for NIR to carry out its mandate as the primary national focal point for all information and documentation in the country and to prepare for it to participate in regional networks such as SADIS and PADIS.
Another objective of the project, although unstated, was to gauge the volume of socioeconomic information being generated on or about Botswana each year.
BOX 1 DEVSIS Solves Information Problems
When we were making efforts to establish SADIS, the Government of Botswana also decided to focus its development efforts on the communal areas; that is, those areas where land tenure and resource management still follow the traditional communal pattern. There was a feeling that the problems of effecting meaningful, self sustaining improvements in the standard of living in the communal areas were among the most intractable Botswana faced. The Government's wanted to tackle these problems by driving for development in these areas but they recognized that their efforts would be hampered if they were unable to take advantage of the wealth of information that already existed. DEVIS - Botswana was the perfect way to con" front their concern.
PROJECT EXPERIENCE AND IMPLEMENTATION
DEVINDEX - Botswana
Soon after the project began, the Project Coordinator, who had actually been involved in the conception and design of the project, had to leave NIR. I had to take over the day-to-day management of the project because identifying another project coordinator would have delayed implementation. My initial role was technical coordinator and I was responsible for abstracting and indexing documents and checking on the quality of cataloging.
DEVSIS-Botswana was conceived as a national project to collect, organize, and disseminate information related to Botswana's economic and social development. Unfortunately, we lost track of this objective somewhat as we implemented the project. For example, throughout the project implementation period, there was no contact with the Government ministries nor with other institutions that had been consulted when the project proposal was prepared. The project, therefore, was generally regarded as an NIR one and not as the collaborative effort that it had been intended to be. We lost the potential national support that had been promised at the initial stage.
I should, however, point out that there was no deliberate decision on the part of NIR to exclude other important partners. We were very much conscious of this fact but the implementation schedule of the project was such that we had to give priority to the production of a printed version of DEVINDEX-Botswana. Because of this production schedule, there was also little time to focus on identifying new information sources for the proposed database. We therefore had to reprocess the documents already available in the library.
When the project started in 1984 nobody knew anything about operating a computer. No software had been identified for the proposed database. The reason for this, however, was that it had been decided that the computer processing of data would be done in PADIS and this had seemed feasible until the implementation of the project. The library was already using its own manual information processing and retrieval system, which was tedious but very appropriate for the collection. The form we used for information processing was simple but adequate for our purposes. However, processing documents for the PADIS database required that we use different and more complicated information processing methodologies designed by PADIS. This caused long delays in the processing of documents. We did not realize until much later that the delays were not being caused by the methodologies we were using but because we were using two methodologies simultaneously. The situation changed when we decided to adapt the PADIS methodologies to suit our internal operations.
PADIS sent one of its indexers/abstractors for three weeks to assist in the manual processing of the documents. His stay was very valuable since we did not have any experience with processing information for a computer-based system. The input sheets were then sent to PADIS for entry into the computer and eventually to IDRC for the production of the first DEVINDEX-Botswana. The ensuing issues of the index were to be a collaborative effort between NIR and PADIS. This arrangement had some practical problems and it was eventually decided that NIR should do the best that it could. After a lot of trial and error, two more indexes were produced. To date, six issues of this index have been produced. In a way, however, this trial and error provided good training.
The information in the six indexes was not selective. Everything on Botswana that had been processed over a particular period was included. Initially, there was no problem with this since the collection was small but, during the production of the sixth issue, it became clear that some criteria were necessary for selecting what should be included in the index. The problem was not only the size of the index but the cost of its eventual distribution. It turned out that the problem was that the index was rather general in terms of subject coverage and this made it difficult for us to target specific audiences.
In retrospect the manual processing of documents should have been followed by the creation of a computerized database at NIR and the actual production of printed indexes. This would have provided a practical training program that would have included everybody in the Documentation Unit and would have helped us avoid the number of problems that we experienced. It was not, however, until 1987 that the first library database was created at NIR.
Research institutions and information centers tend to be subject-specific. There are of course, big libraries in America, England, and Europe that are interested in all the information about a country, but these are very few compared to the subject-specific ones. I decided that the main library's bibliographic database should be structure to enable us to produce specialized indexes and bibliographies according to the focus areas of NIR: agriculture; education; environment; health and nutrition; rural development; and women and gender issues.
This does not, however, mean that there are six different databases: instead, there is one comprehensive database' subdivided by codes. I find this system good because searching can be done on one database or across different subject areas. The actual production of specialized indexes or bibliographies can be subject-specific. This not only makes networking more effective but makes marketing of the products a lot easier. DEVINDEX-Botswana will, however, continue to be produced containing information about Botswana that does not fall within the specified areas. My experience with the specialized databases shows that there will still be some overlap.
The Meetings Database
The main purpose of this database was to collect information on forthcoming meetings, conferences, seminars, and workshops and to disseminate this information to users. The database focuses on meetings taking place outside Botswana because the information about these is available through journals, magazines, and newsletters that NIR receives. One problem we face is that the information often reaches NIR very late and users do not have enough time to register their interest or. more importantly, arrange for funding. The information collected, however, still serves a useful purpose. Because we knew what meetings are taking place, we are able to ask for reports of those meetings. Electronic mail should make it possible for information to be obtained as soon as it becomes available.
Newspaper Clippings Database
The Documentation Unit provides a newspaper clippings service. When the service started, about ten years ago, articles found relevant were clipped and then stored in folders, where they were seldom used. I decided that each newspaper article should be entered into the computer to facilitate immediate and easy access and to make it possible for one article to be used for different purposes.
A good example is that of AIDS, where an analysis of the issue can be made from a variety of angles: AIDS end women; AIDS and Youth; or AIDS in Botswana. Computerizing the newspaper article collection has put more emphasis on easier and quicker retrieval of information. This is actually an easy income-generating activity since a lot of articles are requested through the photocopying service. (See Box 2.)
Mailing List Database
For a long time the library maintained a mailing list of all the institutions with which it had any links. The main purpose of the mailing list, however, was for exchanging information and, particularly, distributing the Institutes' publications. The system used was a manual one, which was effective but very tedious and time-consuming. The only way that this information could be retrieved was by name of institution.
A proper mailing list is a very important management tool that many organizations tend to take for granted. I came to appreciate this only after I had computerized the library database using Micro CDS/ISIS. In its computerized form, it is now possible to retrieve only those institutions or individuals that should receive the NIR's publications list, for instance. It is also possible to retrieve names of institutions and/or individuals that should receive certain publications free of charge or only information about the availability of those publications, so that they can decide to order.
One of our researchers had traveled overseas and desperately needed a list of institutions with which we had exchanged agreements within the area of health. It was possible to provide this information in the shortest time possible because the database is structured in such a way that the computer can pull out first those institutions with whom we have an exchange agreement with and then narrow down the selection to the relevant subject area.
BOX 2 Secondary Information Sources
It is worth noting that a lot of libraries and documentation centers do not seem to appreciate the very important role that newspaper cuttings play as a supplementary source of information' particularly in developing countries where research has a very recent history.
THE RESULTS, IMPACT, AND BENEFITS OF THE PROJECT
Impact of the DEVSIS-Botswana on the Documentation Unit
The Documentation Unit felt the most immediate effect of DEVIS. Before the project, we used a card catalog that was very tedious - from the production of the cards to their management after they had been filed. No longer do we type the cataloging information onto a stencil that then has to be reproduced elsewhere. Of course, the list of possible tasks that always took priority over the production of the cards was endless and all too often the machine was out of order. No longer do we go through the most tedious exercise of filing the cards. Since we were using a controlled vocabulary, we had to update the cards everytime the thesaurus was changed or updated. The automated database relieves us from all of these clerical chores. The information can be easily updated and we can concentrate on providing quality services to our users.
Impact of the DEVSIS-Botswana Project Outside the Institute
The impact of the project outside the NIR Documentation Unit can be judged by the number of requests for assistance with database development that we have been receiving and continue to receive. The following are excerpts from some of these requests:
"In terms of computerization, our center has acquired the necessary hardware and chosen Micro CDS/ISIS software. I believe your documentation center uses this software and therefore offers the best prospects for attachment training. . . The training program that we have in mind is one that can provide: a) an understanding of the database concepts, and expose participants to the skills required for the design and implementation of a database management system; and b) skills in analyzing, designing and implementing computer based information systems." (National Council for Scientific Research, Lusaka, Zambia, 1988)
"Mr. M has recently attended a short course on Improving the Effectiveness of Small Libraries and Information Centres held at the University of Botswana, where among other things he was introduced to CDS/ISIS software and PADIS methodologies. From his report on the course, we are made to understand that the National Institute of Development Research and Documentation of the University of Botswana is one of the best and most successful national information centres within the PADIS. It is in this regard that we would appreciate it if you can arrange for a study visit to enable Mr. M to gain practical experience on the application of various techniques of information handling, storage, retrieval, and dissemination." (The Institute of Finance Management, Dar es Salaam,Tanzania, 1990)
"I would like to take advantage of my contact leave to come and visit your Documentation Centre, and have a first-hand practical experience with your newspaper indexing project." (Mr. E.R.T. Chiware, Periodicals Librarian, University of Zimbabwe Library, 1991)
"The University of Bophuthatswana is in the process of setting up a documentation center. I, therefore, wish to send two professional librarians, to study your setup, especially the organization/processing of materials." (University Librarian, University of Bophuthatswana Library, 1993)2
"Through Professor Heywood I heard about your computerization project and DEVSIS-Botswana, and I am greatly interested in paying you a visit to study your work in this regard." (Director of the National Archives of Namibia, 1991)
The above are only samples of the many requests that we have received. As can be seen from excerpts, however, the requests come from a variety of countries and institutions. This, in my opinion, is a clear indication of the impact that the project has had outside NIR. I would like to point out that, unfortunately, it has not been possible for us to satisfy these requests due to staff constraints. Attachment programs require a lot of time for participants to get the maximum benefit out of them. We have, however, had some people spend one to three months with us learning the processing and management of unpublished literature through the use of micro CDS/ISIS.
ANALYSIS OF LESSONS LEARNED
Project activities should be integrated into the normal activities of the organization. This has implications on the type of organization selected to undertake a project such as DEVSIS-Botswana. I believe that whatever amount of success that NIR may boast about is due to the fact that we did not have to deviate a lot from our normal tasks in order to accommodate the project activities.
If, as was the case with the DEVSIS-Botswana project, immediate results are expected, professional and experienced staff should be employed. In the case of the DEVSIS-Botswana, the aim was not only to obtain immediate results, but also to strengthen the staffing situation.
The schedule that we actually followed when implementing DEVIS-Botswana was neither efficient nor effective. It was dictated more by circumstance than by design. If I was able to start over again, I would definitely follow a schedule that looked more like the following:
1. Employment and orientation of a professional librarian who will be responsible for database management.
2. Training of the professional Librarian for about three months in computers and database management. The latter should include thorough training in abstracting and indexing since these are very important skills for a database manager.
3. Familiarization visit to PADIS by the Project and Technical Coordinators. The Project Coordinator should always be someone who was involved in the initial designing of the project. The Technical Coordinator is responsible for abstracting and indexing of documents.
4. Putting a network in place for purposes of identifying information sources. The sources should not only be for printed documents, but also for information about research activities taking place. This way, we would have ended up with two types of databases - a bibliographic one and a research inventory database. I found that simply tracking hard documents misses out on getting to know who the real information generators are.
5. Collecting documents that will be processed for the project.
6. Employment of support staff including a Data Entry Clerk. It is very important that the latter be employed at the beginning of the project so that data entry is given full-time attention. The mistake that NIR made was to depend on the secretarial staff of the Institute who always had other tasks to take care of and also never really got to learn the library software since they did not use it all the time.
7. Training of the project staff in basic computer skills and PADIS methodologies by PADIS.
8. Processing of documents and entering data in the computer.
9. Publicizing the project, including organizing seminars for potential partners and policy makers and working with the press.
10. At this stage the database manager should be back from training and can start checking on the data entered.
11. Production of printed outputs, which would include not only one index but also some specialized bibliographies. A conscientious document collection exercise should result in a big enough collection to facilitate this. Two research directories would also be produced: one on ongoing and completed research and another on research institutions. The advantage of leaving the production of printed outputs until the very end is that there will be very little editing to be done.
Another lesson to learn is not to allow or encourage the development of bibliographical databases independently of the ones that have been created by the library or documentation center. Work can be subcontracted to individuals to set up these databases but the information should always be transferred to the main database. (See Box 3.)
The NIR also embarked on a research program to build a database on whatever documentation was available on the Basarwa (bushmen). I succeeded in making this part of the existing library database. The good thing that came out of this negotiation is that this information is not only used for the Basarwa research program but is also used for research on the remote area development program that is intended for disadvantaged communities, including the Basarwa/San.
This helps prove that, once information has been collected and processed, it cannot and should not be compartmentalized. It cannot be argued that because a database is developed on environmental issues, it will only be of interest to environmental researchers. For example, I set up a database on energy up for the African Energy Policy Research Network. As I processed documents for this database, I noted the close relationship between energy and environment. Fortunately, I had already negotiated that the database be done as part of the existing library database. I have tried as much as possible to check how often the information from this database is used by other researchers outside the energy field and there is a strong indication that merging the two databases was the best decision.
BOX 3 Lost Information
This mistake was made at NIR: a few databases were created as special projects. Information was entered in computers outside the library and no arrangements made to transfer this information back to the library. The result was that, for a number of reasons, these databases disappeared and a lot of valuable information was therefore lost. This was unfortunate because the program that is used for the development of databases in the Institute has facilities for merging different databases.
CONCLUSIONS AND RECOMMENDATIONS
In conclusion, I would like to recommend the adoption of the Basarwa Database/Bibliography as a model database development project for two reasons. First, the Documentation Unit was fully involved in the project from the beginning - even though it had been conceived elsewhere. Second, this project includes the following "main ingredients" of a database development project.
Pre-Project Consultations Among Potential Project Participants
The idea of the project came from Professor Sidsel Saugestad, Research Facilitator for the Remote Area Development Program (RADP). The NIR Director, various officers at the Ministry of Local Government, Lands and Housing, other individuals interested in the Basarwa issues, and I all consulted on the project. I was involved to ensure that all the technical aspects were taken care of at the beginning. It was very important that the Ministry was involved so that the necessary link between NIR and Government could be established at the very beginning.
Collection of Documents for the Database
A research assistant was employed to go through the NIR library databases and identify Basarwa-related documents so that they could be re-indexed with the potential users of the database in mind. The research assistant also visited other local libraries. Through visits by the research assistant and enquiries from individuals involved in Basarwa research by Professor Saugestad, we came to know of the existence of many documents that we would not otherwise have known about.
We are using a very interesting method for collecting documents. We try and make contacts with researchers who have published widely in the area of the Basarwa studies and ask them to send us lists of their published and unpublished works and the actual publications. We have received a number of these lists and are using them to follow up more documents. The advantage of having these lists is that we are able to check against what has already been collected and what is available.
Provision of Funding for Production of Abstracts
Very often, organizations realize the need for information on specific subjects but never seem to realize the need for providing funding to make this possible. The result is that, through donor funding, databases are developed as special projects that end as soon as funding is exhausted and printed copies of those databases have been produced.
For the Basarwa Database, funding for the production of abstracts was provided. This left me only with the responsibility of checking the technical aspects of the database and editing the abstracts. Some people may wonder why I am emphasizing the inclusion of abstracts in a bibliographic database when they are so demanding to produce. I can only say that it is not until you provide these that you realize just how important they are to your users.
The availability of funding made it possible for us to be selective in our choice of abstracters. Mrs. Janet Hermans, an anthropologist with very keen interest in the Basarwa issues abstracted about 200 documents. Then we hired Shelagh Willet, an anthropologist and retired librarian. She also collects Botswana documents for the Library of Congress and therefore has a lot of experience in collecting grey literature. I therefore recommend that where the situation allows, a specialist in the field should be employed.
Immediate Availability of Documents
One problem we encountered when developing other databases is that documents located in other organizations but processed by us for the database were not retrievable by users at a later date. We could produce a fantastic database but the service falls apart when readers are unable to locate the source documents. All the documents that are being processed for the Basarwa database are already available in the NIR Library. These are being consulted by the researchers in the Basarwa Research Network, university students, government officials, and the general public.
Formalization and Regularization of the Attachment Program
I gave examples above of the requests we receive for assistance with the development of databases using micro CDS/ISIS. I would recommend that an attachment program be formalized and properly and widely publicized so that other people can benefit from our experience. Judging from the number of requests from Botswana and outside, there would definitely be enough participants for the program to run every year.
The program could be structured to provide practical training in information management and micro CDS/ISIS. This would be a lot cheaper for institutions in Africa. At the moment, we only take participants from one organization at a time. This, I feel, is not at all cost-effective. I would recommend that the training be done for groups of participants so that they can share experiences and learn from each another.
Our program does not teach participants the basic concepts of computers and information science; we require that they already have experience in running their libraries or documentation centers. Instead, we expose them to the various information services that we provide. We then work with them so that they get to know the various processes involved in providing those services. Finally, we let them work on their own like the other staff in the Documentation Unit, but with very close supervision. That way they get to experience a lot of practical problems and through the assistance of the staff, find solutions to those problems.
SO, in conclusion, I would say that we faced and overcame many problems in implementing the DEVINDEX-Botswana database project. We have also managed to produce databases that are very important to our users. The lessons we learned in this process have been put to good use as we design and implement additional databases. Although there were a lot of problems during the implementation stages, in retrospect, they were a very good eye-opener for me. Dealing with the problems gave me an understanding of indigenous information management that I would never have had otherwise.
1. Basarwa is a term used in Botswana for a group of peoples known variously as the Bushmen, Khoisan or San of Southern Africa, and for the purpose of this case study, this is the term that I shall use.
2. Bophuthatswana is one of the "states" that was created during the apartheid era in South Africa.