This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. If you re-use all or part of this work, please attribute it to the United Nations Economic Commission for Europe (UNECE), on behalf of the international statistical community.
Table of Contents
1. Across the world, statistical organisations undertake similar activities albeit with variations in the processes that they use. Each of these activities uses and produces similar information (for example, all organisations define populations for their statistical observations, use statistical classifications, create data sets and disseminate information). Although the information used by statistical organisations is at its core the same, all organisations tend to describe it slightly differently (and often in different ways within each organisation).
2. The Generic Statistical Information Model (GSIM) is the first internationally endorsed reference framework for statistical information. It provides a set of standardised, consistently described information objects, which can be used as inputs and outputs in the design and the production of statistics. As a reference framework, GSIM explains significant relationships among the entities involved in statistical production and can be used to guide the development and use of consistent implementation standards or specifications.
3. As a common language to describe the statistical information, GSIM can facilitate communication within and between statistical organisations. It can provide the foundation for in-depth collaboration, standardisation, or, sharing of tools and methods, and thereby, play an important role in modernising, streamlining and, aligning standards and production associated with official statistics at both national and international levels.
4. GSIM is one of the cornerstones for modernising official statistics and moving away from subject matter silos. It is a key element of the strategic vision of the High-Level Group for the Modernisation of Official Statistics (HLG-MOS) and is endorsed by the Conference of European Statisticians
5. The modernisation of statistical production is needed for statistical organisations to remain relevant and flexible in a dynamic and competitive information environment. It is hoped that statistical organisations will adopt and implement GSIM and the common language it provides. However, a model alone cannot transform an organisation or its processes. In order to meet the future needs of statistical organisations, GSIM is designed to allow for innovative approaches to statistical production to the greatest extent possible. It is one of the main foundations of the Common Statistical Production Architecture (CSPA)
, a collaborative initiative to design common and interchangeable services with standard interfaces that support standardisation and modernisation. At the same time, GSIM supports current ways of producing statistics.
6. This paper provides an introduction to GSIM, summarising the key points for a relatively general statistical audience. For more detail, please see the GSIM documents available on the GSIM wiki
7. GSIM provides the information object framework supporting all statistical business processes such as those described in the Generic Statistical Business Process Model (GSBPM)
, giving the information objects agreed names, defining them, specifying their essential properties, and indicating their relationships with other information objects. It does not, however, make assumptions about the standards or technologies used to implement the model.
8. GSIM does not include information objects related to activities within an organisation such as human resources, finance, or legal functions, except to the extent that this information is used directly in statistical production. For more information on these activities see the Generic Activity Model for Statistical Organisations (GAMSO)
9. GSIM is a conceptual model and does not prescribe how the information should be implemented. Organisations can choose existing standards (e.g. SDMX, DDI) for the technical implementation (for more, see section “SDMX, DDI and other standards”).
10. GSIM contains objects which specify information about the real world – “information objects”. Examples include data and metadata (such as statistical classifications) as well as the rules and parameter inputs needed for production processes to run (for example, data editing rules). GSIM identifies around 130 information objects, which are grouped into five top-level groups as in Figure 1.
Figure 1. GSIM Top-level information object groups
11. The five top-level groups are described below. Information objects in the GSIM model are given in italics.
12. Figure 2 shows a simplified view of the information objects identified in GSIM. It gives users examples of the objects that are in each of the five top-level groups.
Figure 2. Simplified view of GSIM information objects
13. Figure 3 shows another view of one part of GSIM. This is a slightly more technical view but still intended to be accessible by a relatively wide audience. Both Figure 2 and Figure 3 can be used as a means for communication with users who are interested in examples of the objects and relationships in GSIM.
Figure 3. Alternate simplified view of GSIM information objects
14. Figure 3 gives an example of GSIM information objects that tell a story about some of the information that is important in a statistical organisation.
“A statistical organisation initiates a Statistical Program. The Statistical Program corresponds to an ongoing activity such as a survey or an output series and has a Statistical Program Cycle (for example, it repeats quarterly or annually).
The Statistical Program Cycle will include a set of Business Processes. The Business Processes consist of several Process Steps which are specified by a Process Design. These Process Designs have Process Input Specifications and Process Output Specifications. The specifications will often be pieces of information that refer to concepts and structures (for example, Statistical Classification, Variable, Population, Data Structure, and Data Set).
If, for example, the Business Process is related to the collection of data, there will be an Information Provider who agrees to provide the statistical organisation with data (via a Provision Agreement). This Provision Agreement specifies an agreed Data Structure and governs the Exchange Channel used for the incoming information. The Exchange Channel could be, in this case, a Questionnaire or an Administrative Register. It will receive the information via a particular mechanism (Protocol) such as an interview or a data file exchange.
The Data Set produced by the Exchange Channel will be stored in a Data Resource and structured by a Data Structure.”
15. More information about the groups and their information objects can be found in the GSIM wiki.
16. It is intended that GSIM may be used by organisations to different degrees. It may be used in some cases only as a model to which organisations refer when communicating internally or with other organisations to clarify the discussion. In other cases, an organisation may choose to implement GSIM as the information model that defines its operating environment. Various scenarios for the use of GSIM are valid, although those organisations that make use of GSIM to its fullest extent may expect to realise the greatest benefits.
Long term benefits
17. GSIM provides a set of standardised information objects, which are the inputs and outputs in the design and production of statistical business processes. By defining information objects common to all statistical production, regardless of subject matter, GSIM enables statistical organisations to rethink how their business could be more efficiently organised.
18. GSIM could be used to direct future investment towards areas of statistical production where the common need is greatest. It could also enable some degree of specialisation within the international statistical community. For example, some organisations could specialise in seasonal adjustment, time series analysis or data validation, and other organisations could take advantage of this expertise.
19. Implementation of GSIM, in combination with GSBPM, will lead to more important advantages. GSIM could:
20. A significant benefit of using GSIM is that it provides a common language to improve communication at different levels:
21. Improving communication will result in a more efficient exchange of data and metadata within and between statistical organisations, and with external users and suppliers.
22. GSIM can be used by organisations to:
23. GSIM has links to several models have been developed under the auspices of the HLG-MOS to support the modernisation of official statistics (e.g. GSBPM, GSIM and CSPA, collectively called the “ModernStats” models with GSIM).
24. GSIM and GSBPM are complementary models for the production and management of statistical information. GSBPM models the statistical business process and identifies the activities undertaken by producers of official statistics that result in information outputs. These activities are broken down into sub-processes, such as “Edit and impute” and “Calculate aggregates”. As shown in Figure 4, GSIM helps describe GSBPM sub-processes by defining the information objects that flow between them, that are created in them, and that are used by them to produce official statistics.
Figure 4. GSIM and GSBPM
25. Greater value will be obtained from GSIM if it is applied in conjunction with GSBPM. Nevertheless, it is possible (although not ideal) to apply one without the other. In the same way that individual statistical business processes do not use all the sub-processes described within GSBPM, it is very unlikely that all information objects in the GSIM will be needed in any specific statistical business process.
26. Good metadata management is essential for the efficient operation of statistical business processes. Metadata are present in every phase of GSBPM, either created, updated or carried forward unchanged from a previous phase. In the context of GSBPM, the emphasis of the over-arching process of metadata management is on the creation, updating, use and reuse of metadata. Metadata management strategies and systems that are developed and maintained in GAMSO activity areas are therefore vital to the operation of GSBPM and are facilitated by GSIM.
27. GSIM can also support a consistent approach to metadata, facilitating the primary role for metadata envisaged in Part A of the Common Metadata Framework "Statistical Metadata in a Corporate Context"
, that is, metadata should uniquely and formally define the content and links between objects and processes in the statistical information system.
28. Describing statistical business processes and their inputs and outputs using the standardised vocabulary of the GSBPM and GSIM supports architecture-management functions by:
29. CSPA uses GSIM as a common reference when defining the information input into, output from and support of business processes. Using GSIM as a common language increases the ability to compare information within and between statistical organisations and hence facilitate the development of harmonised and re-usable services and components.
30. GSIM will help you in effectively managing your organisation’s business architecture by providing a standardised list of information objects used in the statistical production. Mapping actual inputs and outputs of the statistical production onto information objects of the GSIM supports standardisation across subject matter domains.
31. GSIM will help you to improve your communication with colleagues (both locally and internationally). Communication of subject matter between domains is often poor, making the sharing of concepts, variables, and design components difficult without a complex mapping exercise. GSIM can serve as a common language and will ease communication between:
32. GSIM will help you design and understand your processes (and their inputs and outputs) better.
33. For a production cycle, a statistician can design the input and the output, and the process in-between. In GSIM terms, the output and the input can be designed in terms of structures and concepts information objects, and the process in-between can be designed using the business information objects. The structures and concepts objects are provided by subject matter specialists.
34. As seen in Figure 5, if the GSBPM is considered as a frame of reference for statistical production processes, the top level can be considered as equivalent to the statistical production process as a whole. The next level down corresponds to a phase of the statistical production process (for example the “Process” phase of the GSBPM). The third level corresponds to a sub-process (for example sub-process 5.3 of the GSBPM – Review and validate). The fourth level consists of the individual building blocks within the sub-process, such as detecting financial values that might be expressed in thousands rather than units.
35. An important issue for statisticians is the problem of single-use design components, which are often recreated or at least modified for each production cycle. GSIM facilitates the description of inputs and outputs at each level of the GSBPM, following the same pattern thus providing a consistent structure to design statistical processes. It supports the design, specification and implementation of harmonised methods and standard technology to create a generalised statistical production system.
36. Using GSIM will enable producing reusable and flexible process building blocks which can be used by statisticians to produce final products of varying complexity, facilitating the production of a wider variety of products and responding more easily to changing client needs.
37. The use of GSIM, in combination with other ModernStats models, will reduce workloads as many processes can be repurposed and reused. This means less time spent on repetitive work and more time for innovation.
38. In the long term, GSIM, in combination with other ModernStats models, will make statisticians less reliant on information technologists. Due to tools that can be designed and developed to be parametrized for dealing with projects from different domains.
39. Statisticians are very much concerned today about the applicability, usability and stability of their methods and technical solutions. In the “stove-pipe” approach to statistical production, the subject matter specialist is heavily dependent upon the information technologists in the design, build and production of statistical systems.
40. Statisticians will gain greater control over the design of their processes making them more self-supporting in the design and production of their statistics.
41. Production will be based upon more standardised applications that are more robust to change and less vulnerable to changing personnel. An increase in the use of standardised applications implementing standardised methods, which can easily be shared across domains, will enable statisticians to more easily work in different domains.
Figure 5. GSIM information objects
in the context of GSBPM
42. The main concern for information technologists is the duplication of effort due to the “stove-pipe” organisation of statistical production. Unstable and different requirements from these “stove-pipes” lead to tailor-made one-off solutions, whilst a high turnover of Information Technology (IT) staff can result in poorly documented and non-standard applications.
43. The introduction of GSIM both at the national and at the international level can bring short term benefits for IT specialists. GSIM will provide a common language for information technologists to talk to clients and colleagues both locally and internationally. The semantic network of information objects provided by GSIM helps to understand their intrinsic relations and to improve the design of platforms of systems that are more interoperable with each other.
44. At the national level, statisticians will become more self-supporting in the design (see Figure 6) and the production of their statistics reusing and repurposing harmonised components as GSIM, together with other ModernStats models, will enable more flexible and modular production systems. Production will be based upon more standardised applications that are more robust to change and less vulnerable to changing of IT personnel. An increase in the use of standardised applications, which can easily be shared across domains, will enable the IT specialists to more easily work in different domains.
45. The use of GSIM will reduce the workload as many components can be repurposed and reused. This means less repetitive work and more time for innovation.
46. This will free the IT staff to make more robust applications and explore new ways to better meet the changing needs of the statistical organisation and their clients at large. This will include more time for the creation of robust, modular, harmonised, well-documented processes that comply with the requirements of CSPA.
Figure 6. Design your own imputation process
47. At the international level, there will be increased possibilities for co-design and co-development of common components based upon more robust user-requirements from a wider user community. The IT developers will also have access to a larger development community that all speak the same language to describe their statistical information.
48.Using GSIM (especially when it is used with GSBPM together) as a base for standardised information objects can support various management activities covered by GAMSO activity areas:
49. As a reference framework of information objects, GSIM has a complementary relationship with standards
, such as Statistical Data and Metadata eXchange (SDMX) and Data Documentation Initiative (DDI), which are commonly used to represent and exchange statistical data and metadata.
50. The information objects within GSIM are conceptual; no specific physical representation of the information is prescribed. As a simplified illustration, the name of an organisation can be defined as the same concept regardless of whether the information is recorded in a database, in a spreadsheet, in a CSV file, in an XML file or handwritten on a piece of paper.
51. GSIM allows organisations to start with a common language related to the data and metadata used throughout the statistical business process. In this context, GSIM information objects have been mapped to relevant representations in SDMX and DDI.
52. This will help statistical organisations to describe and manage statistical information using a common language while, at a systems level, the information is represented and exchanged in an appropriate and standard technical format.
53. While GSIM information objects can be mapped to SDMX and DDI (and substantial business benefit can be obtained from harnessing these standards), GSIM does not require these standards to be used. Some producers and some users of statistics may decide to use alternative standards for particular purposes. In other cases, producers of statistics may be open to using SDMX and/or DDI but have legacy information systems which are not economical to update for use with these standards.
54. Describing statistical information using GSIM as the common point of reference helps users to identify the relationship between two sets of statistical information which are represented differently from a technical perspective.
55. For example, a statistician may receive some data described in DDI and some described in a locally created format. The statistician can relate both of these to GSIM. The statistician will be able to identify which differences are purely technical and which reflect underlying conceptual differences.
56. Once the nature and extent of the differences can be understood, it often proves straightforward to transform the information into a common technical representation (for example, SDMX or DDI) which allows the content to be integrated and explored. This approach ensures that the results of the technical conversion to a common standard are accurately understood, and are sound, from a conceptual perspective.
57. There are a number of synergies between the use of GSIM as a reference framework and the application of representation standards such as SDMX and DDI. These synergies have been maximised by design.
58. For example, when determining the set of definitions to be used for information objects within GSIM, existing standards and models were harnessed as key reference sources. While none of these existing sources had the same purpose and scope as GSIM – that is a reference framework of information objects spanning the full statistical business process – the development of each entailed analysing and supporting particular needs and scenarios related to particular types of statistical data and metadata.
59. In this way, GSIM benefited from the investment of time in analysis, modelling, testing and refinement when developing these standards and models to their current level of maturity. It also means GSIM does not vary “for no reason” from terms and definitions which are used in existing standards and models. Where it does vary, it is for reasons such as existing relevant standards and models being inconsistent internally, with one another and/or statisticians reporting that alternative terms or definitions are more relevant to their business needs.
60. This paper introduces GSIM to those who work in statistical organisations. It outlines the benefits of the model as well as how the adoption of the model might benefit staff in statistical organisations. The paper also discusses the interaction of GSIM and other frameworks and models such as GSBPM, CSPA, DDI and SDMX.
61. In addition to providing a reference framework for statistical information, GSIM aims to provide ideas and help for modernisation of official statistics. Substantial amount of knowledge and experience from various statistical organisations is behind its development over the years. With more and more statistical organisations using or planning to use GSIM as a part of their modernisation programme, the user base of the model is steadily growing. Exchange of knowledge and lessons learned benefits new users and experienced users alike and feedback from users is integral for the revisions of the model. GSIM was created by official statistics community for official statistics community, all GSIM users are invited to share experiences and thoughts on the GSIM wiki and collectively shape the future development of the model.