I. ORGANIZATION OF THE SEMINAR
1. The Seminar on Application of Information Technology in National Statistical Offices, organized jointly by the secretariat of the Economic and Social Commission for Asia and the Pacific (ESCAP) and the National Statistical Office of the Republic of Korea, was held from 15 to 18 December 1998 at Taejon, Republic of Korea. The Government of the Republic of Korea provided financial assistance for the organization of the Seminar through the Korea-ESCAP Cooperation Fund.
2. The Seminar was attended by 39 representatives of the following 22 members and associate members of ESCAP: Armenia; Australia; Bangladesh; Bhutan; China; Fiji; Hong Kong, China; Indonesia; Kazakhstan; Lao People's Democratic Republic; Macau; Malaysia; Maldives; Mongolia; Myanmar; Nepal; Republic of Korea; Samoa; Sri Lanka; Tonga; Vanuatu; and Viet Nam.
3. The Seminar was also attended by the representatives of the Asian Development Bank (ADB), the Pacific Community (SPC), and the Statistical Institute for Asia and the Pacific (SIAP).
B. Opening of the Seminar
4. The meeting was opened by Mr Young-Dae Yoon, Commissioner of the National Statistical Office of the Republic of Korea. In welcoming the participants, he noted that their attendance demonstrated the importance to governments and organizations of the utilization of modern information technology in producing and disseminating official statistics. The Republic of Korea was showing strong signs of recovery from the recent financial crisis which had affected many countries of the region. The crisis had boosted the demand for statistics but had also impeded their production. Exploiting current technological innovations would help national statistical offices overcome the impact of the crisis by focusing on credibility, customer-orientation, convenience for users, and cooperation with international organizations.
5. The Commissioner described various ways in which his office was utilizing information technology, which accounted for 10 per cent of gross domestic product in the Republic of Korea in 1997. Those included computer-assisted personal interviewing, the launching of a home page, the production of CD-ROMs, and the forthcoming connection of the National Statistical Office database with those of other statistical agencies by means of an on-line network. Such applications would provide the general public with access to comprehensive statistical information in an inexpensive and convenient way. The Commissioner thanked the organizers and resource persons, especially the Australian Bureau of Statistics, for their role in the Seminar. In wishing the participants a pleasant stay, he expressed confidence that their valuable knowledge and experience would make the outcome of the Seminar particularly meaningful and fruitful.
6. The Director of the Statistics Division delivered a message from Mr Adrianus Mooy, the Executive Secretary of ESCAP. The Executive Secretary expressed his deep gratitude to the Government of the Republic of Korea and especially the National Statistical Office for their generous support in making the Seminar possible. He noted that one of the meeting's main objectives was to produce advice that would help even out the widely differing stages of information technology development among national statistical offices in the region. He cited the United Nations Integrated Management Information System as an example of seizing the opportunity to re-engineer an organization's operating procedures in the light of the possibilities offered by information technology. The Executive Secretary also underlined the fact that successful information technology development had to take account of the size, structure, and objectives of the organization; he felt that the Seminar offered national statistical offices a good opportunity to share experiences in that regard.
7. The Executive Secretary noted the potential that Internet technologies held for national statistical offices, as well as the challenge to implement competitive electronic services for a wide international audience. In addition to dissemination, information technology offered national statistical offices the opportunity to collect data electronically from other government departments as well as from private enterprises. He urged participants to be proactive in initiating such electronic data flows, which would help maintain the reputation of national statistical offices as frontrunners in applying modern information technology. The Executive Secretary also drew attention to the recent recommendation of the ESCAP Committee on Statistics that government departments maintain a high level of awareness of the year 2000 (Y2K) problem in computers and persist with remedial action. In wishing the meeting all success, he expressed ESCAP's readiness to entertain the Seminar's proposals for follow-up action.
C. Election of officers
8. The Seminar elected Ms Roszaini binte Omar (Malaysia) as Chairperson, Mr Le Manh Hung (Viet Nam) as Vice-Chairperson, and Mr Shiu Raj (Fiji) as Rapporteur.
D. Adoption of the agenda
9. The Seminar adopted the following agenda:
- Opening of the Seminar.
- Election of officers.
- Adoption of the agenda.
- The status of information technology in national statistical offices.
- Derivation of full benefits from information technology in national statistical offices.
- Networking solutions for national statistical offices .
- Evolving data collection techniques and technologies.
- Work session on the draft recommendations: best practices in applying information technology at national statistical offices.
- Progress in handling the year 2000 problem in national statistical offices.
- Other matters.
- Adoption of the report.
E. Documentation The documents presented at the Seminar are listed in the Annex to the report.
10. At the conclusion of the Seminar, the participants expressed their deep gratitude to the Government and people of the Republic of Korea, and especially to the Commissioner and staff of the National Statistical Office, for the excellent facilities provided and the warm and generous hospitality accorded to them.
II. THE STATUS OF INFORMATION TECHNOLOGY IN NATIONAL STATISTICAL OFFICES
11. The Seminar reviewed the status of information technology (IT) in national statistical offices (NSOs) on the basis of country/area papers presented, and observed the following general trends in developing economies:
- IT had become essential in statistical and administrative work irrespective of the size of statistical office, but coping with its rapid evolution was a challenge to all NSOs.
- New IT was changing data collection, data processing and data dissemination methods radically, with digitization of data taking place at the data source and electronic records from government administration forming a substantial data source for statistics. With the use of administrative records, some advanced countries were able to cut down the number of statistical surveys and also make census taking and production of the results more efficient.
- Basic national telecommunication infrastructures had developed noticeably, but were still far from adequate for the purposes of electronic data collection from remote areas and effective communication and information retrieval from the Internet.
- Mainframe computers were being phased out in many small and medium-size NSOs in favour of open systems running on PCs (personal computers) or client-server environments. In line with that development, off-line batch processing was being used only in a few NSOs that had not managed to upgrade their servers and build networks.
- In some countries, statistical surveys were being digitized at the source by using notebook and hand-held computers to record interview results. However, paper-based data collection was still a feasible choice for many countries, at least until hand-held computers became much more affordable and easier to launch in wide-scale field use.
- The use of administrative records in census and survey planning, and as a direct source of statistics, was increasing.
- Electronic mail had taken off rapidly, not only becoming a major channel for internal correspondence but also quickly establishing itself as a means for transferring data from field locations to statistical centres.
- Most NSOs had established departmental local area networks (LANs) and many already had wide area networks in place, which facilitated information sharing, internal communications and data collection.
- The Internet was greatly affecting the design of all new statistical and administrative systems of NSOs. Web browsers were being considered as a standard interface to statistical data, databases and information, and Internet technologies (TCP/IP) were replacing legacy systems in internal network and communication development.
- The penetration of PCs had increased but the situation remained far from satisfactory, some offices having no dedicated computers for subject-matter statisticians.
- NSOs were increasingly directed by government-wide IT policies, but many of them still lacked organizational IT policies, including standards for hardware and software.
- Powerful PCs and graphic user interface - (GUI-) based development packages were making end-user-developed applications more common.
- The high cost of the state-of-the-art statistical packages had prompted many NSOs to seek cheaper alternatives for their data analysis, including the use of generic databases and spreadsheets, reliance on free or cheap and light statistical packages, and even the development of their own analytical software.
13. The Seminar observed that small statistical offices had all of their data operations running on PCs, which were often connected to a local area network. It recognized that they had limited possibilities to develop applications in-house or to use off-the-shelf applications which were expensive and difficult to set up, such as data warehousing and data mining tools. However, the Seminar found no reason why even the small offices should not strive at using the latest applicable IT to improve the quality of their data and services.
14. Among the common problems that hampered the effective utilization of IT in NSOs, the Seminar observed the following:
- The most significant constraint in all offices was inadequate human resources, both in terms of number and skills. There was generally inadequate blending of statistical and IT skills, and even the presence of the two was not sufficient to guarantee success. They had to be complemented with planning and project management skills without which applications could not be developed.
- Another problem was the very high turnover of IT personnel. Sometimes IT training tended to defeat its purpose by facilitating the transfer of newly-trained IT staff to more lucrative fields, including the private sector. Statistical offices were not protected from the fact that wages in the public sector were usually lower than in the private sector.
- Hardware was still costly for many purposes. For instance, hand-held or notebook computers were prohibitively expensive for wide-scale survey use.
- Similarly, application development was not cheap when the costs of purchasing equipment, software and development tools were taken into account. The lack of organizational software standards and of standard development methodologies, an insistence on using traditional ways of processing data, and general resistance to change were also hampering IT development.
- The absence of trainers with sufficient skills both in statistics and IT and the lack of inancial resources were a real challenge for organizing effective training for NSO staff.
- Basic infrastructure services, especially electricity supply and telecommunications, were poorly developed in some countries, making it very difficult to establish electronic data collection and statistical dissemination systems. NSOs working in those conditions had no choice but to protect their systems and data with devices providing uninterrupted services and additional procedures that added overhead to operation costs.
- In some NSOs, IT staff were overwhelmed by relatively simple user support work because statisticians and administrative staff were unwilling to accept the fact that new technology, particularly the power available in today's desktop PCs and the sharing of resources through networks, required them to pick up new skills.
- The still prevalent mainframe applications were increasingly difficult to maintain as the pool of staff knowledgeable in related computer languages was getting smaller.
- Many NSOs faced problems in the development of client-server systems and in the transfer of data and statistical operations to a new environment.
- Many NSOs had operating environments consisting of many types and makes of computers, operating systems, statistical packages and development tools, which added directly and indirectly to operating and maintenance costs.
- There were shortcomings in statistical database development, with many NSOs lacking centralized systems and integration of related functions.
- Some countries had highly inadequate power generation and grids, which hampered all types of IT applications and required substantial investments in devices to secure uninterrupted services and prevent the loss of data.
- Underdeveloped national information infrastructures deprived many NSOs of the possibility of using the Internet for data collection and dissemination.
- Existing data entry and verification systems were often equipped with inadequate features to ensure the quality of data.
- Data back-up systems were still inadequate in some statistical offices, and tape media in particular were not suited for long storage; those deficiencies had resulted in the loss of valuable census and survey data.
15. In concluding the review, the Seminar recommended that NSOs should identify bottlenecks in statistical information systems and IT management and develop action plans to address the problem areas. It emphasized the need to review non-computerized processes and organizational structures critically in connection with IT development projects. The Seminar heard the experiences of several countries in involving senior management in IT development. The modalities used included the establishment of high-level information resource management committees, computer strategy groups and project steering committees.
III. DERIVATION OF FULL BENEFITS FROM INFORMATION TECHNOLOGY IN NATIONAL STATISTICAL OFFICES
16. The Seminar discussed ways of increasing the benefits from information technology on the basis of the paper "Development of best practices on application of information technology in national statistical offices" (STAT/ITNSO/1), and a presentation made by its author, Michael Page-Hanify of the Australian Bureau of Statistics (ABS).
17. Although featuring IT practices in one of the most advanced statistical offices in the world, the Seminar agreed that document STAT/ITNSO/1 contained useful advice for all national statistical offices to follow, and recommended that the secretariat make it available to all NSOs in the region. However, for many less advanced and small statistical offices, some practices were appropriate as long-term goals rather than as advice with immediate applicability. The Seminar agreed that the Australian Bureau of Statistics was in a fortunate position to have a highly developed information culture which appreciated the role of IT as a facilitator of information exchange and as a tool for making work more interesting. The ABS staff had reached a level where they were comfortable with the necessity of using applications personally. These factors brought job satisfaction to staff and, along with the reasonably high remuneration level of IT professionals relative to the private sector, meant that the turnover of IT staff was smaller than in most less developed NSOs. The Seminar recognized that IT development in small NSOs could be shattered by the departure of a key person, which was clearly not a danger at ABS. The Seminar agreed that all those differences required adjustments in the best practices presented and invited NSOs and the ESCAP secretariat to develop the set of best practices further to enhance its effectiveness in all circumstances.
18. The Seminar agreed that by adopting best practices on application of IT, which had worked well elsewhere, NSOs could avoid mistakes and wastage caused by piecemeal development. Best practices were more effective if written down and communicated to everyone within the NSO and to external stakeholders with an interest in the NSO's effective operation. The Seminar noted that best practices were neither a vehicle providing advice on applying particular technologies (that were evolving rapidly) nor a straitjacket tying an NSO to a particular IT path irrespective of its size, but an IT management tool with long-term applicability and a contributor to achieving organizational goals.
IV. NETWORKING SOLUTIONS FOR NATIONAL STATISTICAL OFFICES
19. The Seminar considered the item on the basis of document STAT/ITNSO/2, "A study on the Design Proposal for Unified Statistical Information System using Metadata Modeling", prepared by the National Statistical Office of the Republic of Korea, and a presentation by a representative of IBM Korea entitled "Business Intelligence: Turning Information into Insight".
20. Paper STAT/ITNSO/2 described how the National Statistical Office (NSO) and other statistical agencies in the Republic of Korea had started to disseminate data via the Internet, using independent dissemination systems based on their own IT platforms. That situation caused problems for users such as a lack of common access interface, poor interconnectivity between systems, and no common metadata standards. The document proposed the creation of common definitions and terminology for a unified statistical information system, which would require cooperative work between the NSO and other statistical agencies. Appropriate design principles and considerations and common standards in representing metadata would help create a mechanism for interconnecting different agencies systems. The paper described the advantages and common architectural features of the proposed relational model for a unified statistical database including a unified metadata repository.
21. The presentation by the IBM representative traced the process of turning data into information for decision-making, and featured data warehousing and data mining, which could be described as extracting previously unknown, comprehensible and actionable information from large databases and using it to make crucial business decisions. Data mining was an iterative process and could involve verification of a hypothesis or the discovery of hidden relationships within large datasets. Various data mining techniques and applications were described. The presentation highlighted a case study of the analysis of population and housing census data in the Republic of Korea for 1995, using clustering algorithms.
22. The Seminar was informed that data mining differed from conventional statistical analytical techniques chiefly in the ability to handle very large databases, the use of artificial intelligence-based mathematical algorithms as well as statistical techniques, and better visualization and interpretation properties. It was clarified that data warehousing was not a prerequisite for data mining, but made it considerably easier and more productive. Concerns were expressed about the cost and feasibility of data mining techniques in many of the countries represented at the meeting. The Seminar noted that to obtain full benefits from data mining, several datasets should be utilized, but that implied a commonality of definitions across statistical collections (similar to that being proposed in document STAT/ITNSO/2) which would usually take time and effort to realize.
V. EVOLVING DATA COLLECTION TECHNIQUES AND TECHNOLOGIES
23. The Seminar based its discussions on the agenda item on two papers: "Applications of information technology to tasks of Korea National Statistical Office (STAT/ITNSO/3) by Jung-Jin Lee, Department of Statistics, Soongsil University and Won-Hwan Park, National Statistical Office, Republic of Korea, and Information technologies as applied to the Employment survey in Hong Kong, China (STAT/ITNSO/4) by Anthony Mak Shiu-kwai, Census and Statistics Department, Hong Kong, China.
24. The Seminar heard that the rapid evolution of information technology had changed the ways of conducting statistical surveys in the Republic of Korea significantly. For instance, the GIS (geographic information systems) component, added in 1995, had enhanced the presentation and interpretation of census and survey results. The currently used maps were based on raster graphics, but new vectorized maps were expected to become available in 2000. Computer-assisted interview methods were also improving, and interviewers of the 1998 Survey of economically active population were equipped with notebook computers. The National Statistical Office had developed its own computer-assisted personal interviewing (CAPI) software by using CLIPPER since BLAISE software, which it had initially planned to use, could not conveniently handle Korean characters and the transformation of solar and lunar calendars. The NSO planned to introduce computer-assisted telephone interviewing (CATI) in surveys in 2000. Paper STAT/ITNSO/3 also covered computer-assisted data input, electronic data exchange, and data enumeration and analysis.
25. Document STAT/ITNSO/4 noted that CATI had been initially introduced for the quarterly employment survey in Hong Kong in 1990, and currently covered some 40 per cent of establishments. The paper described the configuration and main features of the new CATI system to be implemented in early 1999. Enhancements included automatic scheduling and telephone dialing, data entry with on-line validation, on-line monitoring by supervisors and interviewer/supervisor intercommunication, management and quality control. New functions covered the transfer of difficult cases to the supervisor, questionnaire despatch and return by fax, bar code scanning, interactive voice responses, and computer-assisted coding. Against a background of increases in the use of pagers and mobile phones, the popular use of fax, and pressures to shorten release times and reduce respondent burden, the new system was expected inter alia to result in productivity gains, higher response rates, better control of staff performance, improvements in data quality, and cost savings.
26. The Seminar heard that one data analysis package mentioned in the Korean paper, ISP, had been developed partly in response to the high cost of the SAS package and the expense of customizing it; moreover, SAS was not commonly used for surveys with complex sampling design because it did not compute appropriate sampling error. The point was made that NSOs sometimes found it difficult to recruit university graduates who were familiar with the software packages actually used in their offices.
27. The Seminar noted that CATI was particularly suited to the Hong Kong employment survey because of the size and frequency of the survey, the small number of questions, and the fact that answers were numerical and non-sensitive; the technique might not be so advantageous in other circumstances, especially in view of the cost of the equipment. In Hong Kong, respondents often preferred CATI since it took less time than a field visit; data quality did not appear to suffer. The Seminar noted that although receiving information by fax was not a major technological advancement, Hong Kong was offering it as an alternative in the employment survey as a convenience to its business respondents. The use of the Internet as a collection mechanism had been considered at the design stage and would be under regular review. The Seminar heard that for field visits, both Hong Kong and the Republic of Korea used notebook computers in order to perform editing in the field, although it was acknowledged that much cheaper hand-held computers were a good choice in other situations requiring data entry only.
VI. WORK SESSION ON THE DRAFT RECOMMENDATIONS: BEST PRACTICES IN APPLYING INFORMATION TECHNOLOGY AT NATIONAL STATISTICAL OFFICES
28. The Seminar recalled its earlier discussions and decided on the following set of recommendations, which were framed to be considered jointly by both senior management and senior IT staff of the organization. The Seminar recognized that, depending on the current state of IT development of the respective NSOs, the timeframe in which implementation of the recommendations could commence would vary. Finally, the way the recommendations would be adopted depends largely on the role of the IT department has within the organization, but as a minimum, the IT department should formulate a set of "best practice" policies applicable to their organization.
- A high-level committee consisting of senior management and senior IT staff should be formed to consider matters of strategic significance concerning data and information management and related policy.
- An IT strategic plan should be developed by senior IT staff which takes into account the overall mission and goals of the organization. This document should be used as a firm guideline by the IT department. It would enable the NSO to better integrate data across statistical areas, reduce respondent burden, ensure less costly means of data collection, formulate financial requirements, and so on. The plan would be endorsed by the committee formed under recommendation (a).
- Official IT standards on the development of application systems need to be established.
- The NSOs should streamline the usage of software, reduce the different versions of existing software being used and put in place a policy for the acquisition of new software and equipment. It is recognized that financial and/or logistical constraints may not permit immediate implementation of the recommendation.
- A disaster recovery plan should be developed; the plan should formulate a policy on the procedures needed to ensure a timely resumption of services in the event of a major break down in the IT environment.
- Strategic relationships should be established with major data providers (e.g., other government agencies) to coordinate the data collection process by applying agreed standards.
- The NSOs should play a more active role in defining the terms of reference of technical cooperation, ensuring sustainable skills transfer with objectives in line with the organization's policies.
- The training of subject-matter statisticians in IT should be initiated and enhanced so that more effective processing is done at the workstation level. The training should be focused and adjusted to the tasks and responsibilities at the various levels.
- The NSOs should implement a plan to retain IT staff by way of a carefully developed training programme geared to upgrade skills, combined with an active programme aimed at better job satisfaction.
- NSOs should follow the existing trend to implement LANs so as to share IT resources as efficiently as possible and also enable more advanced data sharing.
- NSOs should pay greater attention to security requirements for data, which will continue to increase with extra accessibility. They should note that to secure an internal network having outside electronic access a firewall, audited externally where possible, is the only reliable method of avoiding unwanted intrusions and possible introduction of viruses. For smaller NSOs where the cost of a firewall is not practical then a dedicated "stand alone" PC could provide a viable alternative.
- Efficient electronic filing of documents, backup and archiving procedures should be implemented using modern technology.
- Data holdings within the agency should be identified and documented and metadata developed with standard classifications to ensure that efficient data management is practised with prompt information delivery. This will assist commonality of definitions across the various collections.
- An inventory should be made of all hardware and associated software, and of individuals' access to software. This will provide for a smooth upgrade of software as well as keeping users fully informed.
- The possibilities should be explored of organizing electronic communication within the NSOs by implementing an electronic network. This will enable efficient communication and would also help initiate electronic data collection.
- Emerging trends in data capture techniques, such as the use of mobile computers and electronic communication, should be considered when developing new systems or redeveloping existing systems.
- The possibility of organizing electronic data dissemination via the use of advanced IT should be explored, including CD-ROMs, static or dynamic web pages etc. The approach would depend on the volume of data to be disseminated.
- When providing IT services, the focus should be on user requirements with an easy - to - use solution which is stable, reliable and secure. IT staff should be closely involved with clients and wherever possible, should be located in the user area during the development phase to assist them in familiarizing themselves with the organization's business.
- Levels of service should be agreed on with internal stakeholders in order to improve response to users. Service could cover timeliness, outages, response times, and general support.
- For larger NSOs, a permanent help desk should be established to assist in providing a better service for user queries. The service would be facilitated by the formation of a solutions database that the help desk staff populate with the resolution of queries.
- A local network of IT professionals should be developed for cooperation and exchange of views on current IT issues.
29. The Seminar also recommended that ESCAP explore the mobilization of funds from various donor agencies to provide IT infrastructure for NSOs in need. It additionally recommended that similar seminars be organized with IT professionals within the region more regularly. ESCAP was urged continually to advocate efficient utilization of IT in NSOs and to encourage developed countries to accelerate their help to developing countries.
VII. PROGRESS IN HANDLING THE YEAR 2000 PROBLEM IN NATIONAL STATISTICAL OFFICES
30. The Seminar considered the progress in handling the year 2000 problem in national statistical offices on the basis of country papers introduced earlier under agenda item four and of oral reports given by the participants.
31. Several participants had attended the SIAP/ESCAP Workshop on the Year 2000 Problem in Computers and Strategic Issues for National Statistical Offices at Bangkok in June 1998 and reported on follow-up actions to that Workshop. The Seminar noted that the Workshop and its recommendations (which had been made available in hard copy and on the ESCAP Web site) had been useful for NSOs in increasing awareness about and expediting action on the Y2K problem. It observed that the general preparedness level had also improved significantly in most NSOs. In particular, NSOs had made inventories of their hardware, software and systems containing embedded chips, which allowed them to estimate how affected they were by the problem. However, a significant proportion of the necessary remediation work was only about to be started in many of the affected countries.
32. The Seminar heard that many NSOs, including Bangladesh, Indonesia, Malaysia, Myanmar and Sri Lanka, were in the process of phasing out mainframes that could not be made compliant at all or for which remediation was not feasible. Macau had completed that migration earlier in 1998. While client-server systems, including powerful PC servers, were natural replacements for those mainframes, the Seminar cautioned that such development and the migration of data and statistical operations usually took much longer than originally planned, and urged such NSOs to expedite related procurement and application development by using all available means. The Seminar recalled that testing was the most time-consuming part of system rectification and also that systems stated to be compliant had to be tested.
33. While many small NSOs were protected by the fact that they had fairly new stocks of stand-alone PCs with few date-dependent applications, the Seminar however cautioned them to pay attention to the compliance of data suppliers and government agencies that they depended on. For PC-based environments, the upgrade of motherboards only was a fairly commonly used measure to lower replacement expenses.
34. The Seminar noted that Y2K preparedness was satisfactory in Australia and Hong Kong, which had started remediation work fairly early and were progressing on or ahead of the established schedules to achieve compliance; in both places, compliance information and internal guidelines were provided on the World Wide Web. It also noted that the Asian Development Bank and the Secretariat of the Pacific Community were planning to test soon the compliance of their smart building systems by advancing all clocks to the change of the century.
35. The Seminar noted that most countries had established committees to coordinate and monitor the resolution of the Y2K problem at national and sometimes also at the local government level. The Seminar was told that many governments had approved additional funding for Y2K projects on a priority basis, while some countries had to cope with the problem within their normal budget frames. While the compliance of critical systems had to be the priority, the Seminar reminded governments to ensure that valuable old data holdings with the year coded with two digits were protected from corruption, now while still archived and later when read into compliant systems.
VIII. OTHER MATTERS
36. Under the item the Seminar heard a presentation by Ms Dalisay Maligalig of the Asian Development Bank of the ADB's Statistical Database System (SDBS). The SDBS, a fundamental element in the Bank's statistical information system, had been redesigned on a client-server platform in 1996 using the Bank's standards for hardware and software. The presentation covered the challenges in designing the SDBS structure, including the storage of data from different sources with different data attributes and the construction of a well-defined index key, and improved screen displays and database response time. The SDBS featured an on-line help facility, on-line query and report generation modules. Future database developments included a modification to the treatment of external databases, which were currently retrieved into the SDBS, and front-end development using MS Excel.
37. In the ensuing discussion, the following points were among those made:
- although Excel could have difficulty handling a large number of records depending on the computing environment, ADB users rarely wished to deal with that many at one time;
- validation was carried out by running business rules and, checking suspect data with the source;
- the content of the SDBS was determined by the Bank's Interdepartmental Committee on Statistical Data Policy;
- 65 per cent of the data holdings were currently collected in electronic form;
- the Bank planned to launch an electronic questionnaire by 2000 in HTML format, which it hoped would be convenient for respondents although technically challenging for the Bank;
- the ADB had outsourced only the programming of the SDBS;
- from among multiple sources, the Bank chose to publish, where available, data from the organization considered to be the primary national source for that series;
- the Bank had decided to refrain from selling data from the SDBS, which in any event would require consultation with other agencies.
38. The Seminar noted the efforts being made by intergovernmental agencies to reduce the response burden on countries in the supply of data, and heard with interest that Internet technologies were changing the way that agencies gathered data and exchanged them among themselves.
IX. ADOPTION OF THE REPORT
39. The Seminar adopted this report in its draft version on 18 December 1998.
Annex: LIST OF DOCUMENTS
||Title of document
Relevant agenda item(s)
||Annotated provisional agenda
||Development of best practices in application of information technology in national statistical offices (Paper by Mike Page-Hanify, Australian Bureau of Statistics)
||A study on the design proposal for unified statistical information system using metadata modeling (Paper by Ms Jung-im Ahn, National Statistical Office, Republic of Korea)
||Applications of information technology to tasks of Korea National Statistical Office (Paper by Jung-Jin Lee, Department of Statistics, Soongsil University and Won-Hwan Park, National Statistial Office, Republic of Korea)
||Information technologies as applied to the Employment survey in Hong Kong, China (Paper by Anthony MAK Shiu-kwai, Census and Statistics Depeartment, Hong Kong, China)
- STAT/ITNSO/ Armenia
- STAT/ITNSO/ Bangladesh
- STAT/ITNSO/ Bhutan
- STAT/ITNSO/ China
- STAT/ITNSO/ Hong Kong, China
- STAT/ITNSO/ Indonesia
- STAT/ITNSO/ Kazakhstan
- STAT/ITNSO/ Lao People's Democratic Republic
- STAT/ITNSO/ Macau
- STAT/ITNSO/ Malaysia
- STAT/ITNSO/ Maldives
- STAT/ITNSO/ Mongolia
- STAT/ITNSO/ Myanmar
- STAT/ITNSO/ Nepal
- STAT/ITNSO/ Republic of Korea
- STAT/ITNSO/ Samoa
- STAT/ITNSO/ Sri Lanka
- STAT/ITNSO/ Tonga
- STAT/ITNSO/ Vanuatu
- STAT/ITNSO/ Viet Nam
|4, 6, 7, 8, 9