General analysis of standards other than methodological ones
ESSnet Standprep Project
Work Package 2: Processes for the Adoption and Implementation of Standards
Table of Contents
Glossary 2
Executive Summary 2
Methodological Standards and others 4
2.1.1 Process for the Adoption and Implementation of Standards in the CEN Standardisation Process 34
2.1.2 The ISO Standardisation Process 37
2.1.3 Processes for the Adoption and Implementation of Standards: the Planned ONS Harmonisation Process 41
2.1.4 DESTATIS Process (Paper outstanding)............................................................41
Processes for the Adoption and Implementation of Standards: the key elements of the standardisation process for the ESS. 47
3.1 Development of a Common IT Package in the ESS - The Demetra Experience 59
3.2 ARGUS as a De Facto Standard for Disclosure Control 63
3.3 Using XBRL in a Statistical Context – the case of the Dutch Taxonomy Project 64
3.4 SDMX (Paper outstanding)...................................................................................66
CORE ESSnet: Principal Standardisation Efforts 67
Glossary
|ESSnet Standprep Work Package 2 – Glossary |
|ARGUS |Disclosure control software package |
|BT |Technical Bureau |
|BV4 |Seasonal adjustment method |
|CEN |European Committee for Standardisation |
|CENEX |Centres and Networks of Excellence |
|CORA |Common Reference Architecture |
|CORE |Common Reference Environment |
|CRM |Common Reference Model |
|Dainties |Seasonal adjustment method |
|DDI-L |Data Document Initiative-Lifecycle |
|EN |European Standard |
|ESS |European Statistical System |
|GSS |UK Government Statistical Service |
|GSBPM |Generic Statistical Business Process Model |
|GSIM |Generic Statistical Information Model |
|HCSO |Hungary Central Statistical Office |
|HLG-BAS |High Level Group for Strategic Directions in Business Architecture in Statistics |
|HMI |Human-Machine Interfaces |
|ICS |International Classification for Standards |
|INSEE |Institut National de la Statistique et des Etudes Economiques |
|ISO |International Standards Organisation |
|ITIL |Information Technology Infrastructure Library |
|MOF |Meta Object Facility |
|NACE |Classification of Economic Activities in the European Community |
|NSHG |National Statistics Harmonisation Group |
|NSI |National Statistical Institute |
|OCMIMF |Operationalise a Common Metadata/Information Management Framework |
|OECD |Organisation for Economic Co-operation and Development |
|OGD |UK Other Government Department |
|PRINCE2 |Process-based method for effective project management. |
|QMHT |Quality Methods and Harmonisation Tool |
|SDMX |Statistical Data and Metadata eXchange |
|SPSC |Statistical Policy and Standards Committee |
|SRM |Semantic Reference Model |
|StandPrep |Preparation of Standardisation |
|TRAMO/SEATS |Seasonal adjustment method |
|UNECE |United Nations Economic Commission for Europe |
|UML |Unified Modelling Language |
|X11-ARIMA |Seasonal adjustment method |
|X12-ARIMA |Seasonal adjustment method |
|XBRL |eXtensible Business Reporting Language |
|XML |Extensible Markup Language |
Executive Summary
To be completed after the conference by including issues and themes raised during the discussions.
Methodological Standards and others
Table of contents
Identification and classification of ISO standards..................................................5
1. Classification according to the definition of statistical standard....................7
A. ISO standard is a document ...8
B. The recognised body where the consensus was established, and the standard was approved 8
C. The type of common and repeated use 9
D. Rules, guidelines or characteristics 9
E. For activities or related results 9
F. The optimum degree of order in a given context 11
G. Scientific and cultural environment 11
2. For which activity/product can we define standards?.....................................................................................................11
2.1. What kind of activities? 11
2.2. What kind of activity fields? 13
2.3. What kind of results? 13
2.4. Could IT tools be standards in statistics? 14
2.5. The decomposition of statistical activities 16
3. Coherence in standardisation........................................................................18
4. Classification according to the general business statistical architecture......18
Annex 2: Standards of statistics - topics and examples........................................23
Annex 3: Notes on the Generic Statistical Information Model (GSIM)..............27
The grouping and classification of standards can be based on two main aspects:
• the definition of standard and its characteristics; and
• the general business statistical architecture.
The first can give the scope and strength of standard. The other gives the activity field, subject of standard.
The working group accepted the concept of ISO standard as the concept of the statistical standard. At the classification of standards we can try to adapt the ISO classification as well.
Identification and classification of ISO standards
The ISO standards are identified by a prefix, a serial number and the year of publication. For example:
• ISO 860:2007 Terminology work -- Principles and methods
• ISO 10241-1:2011 Terminological entries in standards -- Part 1: General requirements and examples of presentation
• ISO/DIS 10241-2.2 Terminological entries in standards -- Part 2: Adoption of standardized terminological entries
• ISO/WD TS 11669 Translation projects -- General guidance
The prefix identifies the body, the organisation approving the standards (like ISO or ISO/IEC, ISO/DIS, ISO/OECD if the standard is a common standard of the given bodies).
EX: SDMX is a technical specification (i.e, a specific kind of normative document, not a proper standard) and the ID code is: ISO/TS 17369:2005
If the document is other than the main document of a standard, than the prefix shows the type of document (e.g. amendment: Amd, technical corrigendum: Cor, technical specification: TS, technical report: TR, etc.).
The prefix is followed by a serial number which may include a part number, separated by a hyphen from the main number. The serial number of a published standard is followed by the year of publication separated from the serial number by a colon (only if the standard had been published).
The ISO standards are grouped according to the following aspects:
• By ICS, classified by subject in accordance with the International Classification for Standards and
• By the ISO technical committee responsible for the preparation and/or maintenance of the standards.
• By the state of standard (published standard, standard under development, withdrawn standard, deleted standard) and the deeper stage of standardisation process
• By language
From the grouping criteria the subject of the standard is an important aspect for us. During the ESSnet STANDPREP work the most discussed topic was the possible subjects of statistical standards.
The classification of ISO standards by subject are managed by the International Organisation for Standardisation.
“The ICS (International Classification for Standards) is intended to serve as a structure for catalogues of international, regional and national standards and other normative documents, and as a basis for standing-order systems for international, regional and national standards. It may also be used for classifying standards and normative documents in databases, libraries, etc.” [1]
“The ICS is a hierarchical classification which consists of three levels.
• Level 1 covers 40 fields of activity in standardisation, e.g. road vehicle engineering, agriculture, metallurgy. Each field has a two-digit notation, e.g. 43 ROAD VEHICLE ENGINEERING
• The fields are subdivided into 392 groups (level 2). The notation of a group consists of the field notation and a three-digit group number, separated by a point, e.g. 43.040 Road vehicle systems
• 144 of the 392 groups are further divided into 909 sub-groups (level 3). The notation of a sub-group consists of the group notation and a two-digit number, separated by a point, e.g. 43.040.20 Lighting, signalling and warning devices”
In the subgroups always there is an “other” category with the code 99.
One standard can be sorted into more than one subgroup.
The production of statistics is not in the above classification as a subject. There are only some related subjects:
03. SERVICES. COMPANY ORGANISATION, MANAGEMENT AND QUALITY. ADMINISTRATION. TRANSPORT. SOCIOLOGY
03.120 Quality
03.120.30 Application of statistical methods
and
07 MATHEMATICS. NATURAL SCIENCES
07.020 Mathematics
*Application of statistical methods in quality assurance, see 03.120.30
and
35 INFORMATION TECHNOLOGY. OFFICE MACHINES
35.240 Applications of information technology
35.240.01 Application of information technology in general
In classes 03.120 and 07.020, the statistics mentioned relate to sampling methods used in quality control. This does not correspond to the activity field of NSIs, who produce official statistics.
Standard ISO 20252 was established by the Technical Committee ISO/TC 225, Market Studies, opinion and social surveys, and the reference is: ICS 01.040.03; 03.020
Summarizing the classification of ISO standards an example is the description of SDMX standard:
ISO/TS 17369:2005 Statistical data and metadata exchange (SDMX)
Stage: 90.92 | TC 154 ICS: 35.240.01 where
ISO/TS technical specification approved by ISO
2005 is the year of publication
Stage: 90.92 repeat an earlier phase
TC 154 the standard is maintained by the “Processes, data elements and documents in commerce, industry and administration” technical committee
ICS: 35.240.01 the standard belongs to the “Application of information technology in general “subject subgroup
When the grouping of statistical standards is cleared out by the ESS, then the ESS could submit the demand for modifying the structure to the ICS, because the structure is applicable “to all international, regional and national normative documents, such as standards, technical reports, standardized profiles, technical specifications, technical regulations, guides, codes of practice, technology trends assessments, etc. and the drafts of such documents” so the statistical standards as well.
Classification according to the definition of statistical standard
As explained in WP1, a standard is a document, established by consensus and approved by a recognised body that provides, for common and repeated use, rules, guidelines or characteristics for activities or their results, aimed at the achievement of the optimum degree of order in a given context. Note: standards should be based on the consolidated results of science, technology and experience, and aimed at the promotion of optimum community benefits.
In the definition several features of a standard are identified:
A standard is
A. a document,
B. established by consensus and approved by a recognised body( if the consensus does not relate to the member of the body, this should be separated in two)
C. that provides, for common and repeated use,
D. rules, guidelines or characteristics
E. for activities or their results,
F. aimed at the achievement of the optimum degree of order in a given context.
G. Note: standards should be based on the consolidated results of science, technology and experience, and aimed at the promotion of optimum community benefits.
This means, that we can speak of standards if each of the 7 above criteria are true, according to mathematical logic.
Some of the criteria can be true in different ways, and this may lead to different categories of standards.
ISO standard is a document
The ISO glossary shows the document can be of different types, such as the main document of the standard, an amendment, a technical specification, or a user guide.
The recognised body where the consensus was established, and the standard was approved
This category refers to the bodies that the statistical community accepts as a recognised body.
It is beyond the responsibility of ESSnet to decide about such an issue. If we focus more specifically on the ESS, we can agree that:
• ESS defines a community: Eurostat, the set of all NSIs and other statistical authorities, as defined by the regulation 223/2009 (statistical law).
• within this community, it would make sense to retain a single “recognised body” that would be competent for this community - i.e. ESS. There could of course be other bodies but for other communities. For instance, at the national level, Destatis is the recognised body for the standardisation of the German system of official statistics (Destatis and the statistical offices of the Länder). In Hungary, National Statistical Council (known as OST), set standards for the whole national statistical service.
• It makes sense as well to consider that the cultural framework is the same for the community at a given date.
Beyond the ESS, the Organisation for Economic Co-operation and Development (OECD), the United Nations Economic Commission for Europe (UNECE) or the UN Statistical Commission and others (IMF, ILO…) could be considered as standardisation bodies in the future.
In the ISO referencing method the “prefix” can be created by the above A and B features.
Let us denote the ISO identification number in a systematic way as follows:
{Y} {serial number}:{year} - if the standard main document is approved by one body
{Y1}/{Y2} {serial number}:{year} – if the standard main document is approved by two bodies
{Y} {serial number}-{part id.}:{year} - when the standard has more separate parts
{Y} {X} {serial number}:{year} - if the document is other than the main document of the standard and is a specific document type
Then, {Y}, {Y1}, {Y2} is the code of approving body as defined in paragraph B above: Y, Y1, Y2 belong to the list (ISO, IEC, OECD…)
{X} is the code of document type: X can be either blank (for a real standard) or belong to the list (TS, TR, Amd., Cor., Guide[2]…)
{serial number} is the identifier number of the standard without content
{part id.} is the identifier of the part of the standard
{year} is the year of publication
The type of common and repeated use
Beside the approving bodies this criteria can characterize the extension of standard like
• international/world standard
• European standard
• national standard, or other.
However, this is overlapping the definition of the standardisation body. One body is expected to register and manage standards of a single geographical level. Hence, this criteria will not be considered further.
Rules, guidelines or characteristics
The standard has to contain any, some or all of the following: rules, guidelines or characteristics.
For activities or related results
In work package 1 of the grant, the methodological standards have been analysed. If the statistical scope is widened, it is reasonable to consider the business architecture of statistics.
An example of the statistical business architecture is the Bo Sundgren description about the statistical data and metadata model. This model regards the part of the statistical system not only the core of statistical activities but the organisation, the IT environment and the research.
In this case, item E of the definition can be the basis of typology. The standards referring to the results of activities can be differentiated by this aspect if they refer to processes/activities or products/results of activities. Another crosscutting aspect can be the area of the activity or product. In business architecture, it can be differentiated into:
• the core of statistical activities aimed to deliver statistical products,
• IT support
• operation of statistical organisation
• research, training and supporting activities.
This topic is detailed in the following chapters.
Figure 1: Overview of data/metadata resources and flows in a statistical organisation.[3]
[pic]
The optimum degree of order in a given context
The aim of the standard is to achieve the optimum degree of order in a given context. If we take the example of “harmonised standards” in the EU, the aim is to define standards for the goods that are exchanged in the EU internal market, with the appropriate security for the consumer, without overly increasing the cost of products exchanged. Another example would be to consider that the standard should not lead to conditions that cannot be met. For instance, if a standard defines the conditions for an opinions surveys conducted by private companies, considering that random sampling is a rule and that another rule is the response rate should be higher than 60%, would lead to a quality that cannot be achieved in practice. Another example is the definition of standards for products of different types of quality (and in general different levels of cost). Quite typically, there are standards for various types of locks, differentiated in some countries by stars (*,**, ***) or by letters. Then this standard does not aim at the unconstrained optimum degree of security (which would lead to very expensive locks) but offers several security grades that give more choice to the customer.
This attribute can serve as criteria either for the approving or the quality evaluation process.
Scientific and cultural environment
Standards have to be based on the consolidated results of science, technology and experience, and aimed at the promotion of optimum community benefits. The changes in technology, science and experience occur over time. This is why standards may be revised, and why the reference year is mentioned in the identification of the standard.
For which activity/product can we define standards?
Since the production of statistics is considered to be a service activity, standards in statistics will have to be “process standards”.
1 What kind of activities?
The statistical system can be differentiated into four clusters of the overall process representing statistical offices:
1. core of statistical activities aimed to deliver statistical products
2. IT activity supporting statistics
3. operating and organisational aspects of statistical production
4. statistical research and training activities
From this point of view each statistical activity cannot be performed in practice without other types of activities:
• statistical activity demand IT support
• human and financial resources
• support the statistical activity with the new results of science and statistical practice and related training.
Therefore we can imagine the creation of standards for any cluster of the set of all activities in a statistical office.
If we consider the core of statistical activities, the graph associated to the Generic Statistical Business Process Model (GSBPM) (see Annex 1), we can envisage the definition of activities along one major dimension: the sub-processes or phases that are considered in this sequence, from 1.1 to 9.3.
For example, the set of sub-processes 1.1 to 1.4 would correspond more or less to the definition of statistics as agreed in the EU regulations.
Sub-process 5.2 would correspond to the definition of classifications.
Sub-process 2.4 in Business statistics would correspond to the definition of statistical units.
It has to be noted that in the GSBPM, the sub-processes consider together several dimensions:
• work done by humans in the statistical domain
• work done by humans in the IT domain. Traditionally, we could distinguish design/production/maintenance. Another dimension of analysis could be the distinction between Human-Machine Interfaces (HMI) and fully automated actions.
• automated work done by IT systems.
We would end with the following taxonomy:
| |Design |Maintenance |Production |
|HMI |HD |HM |HP |
|Automated |AD |AM |AP |
Since the standardisation aims to all activities in the overall production process, these IT activities could/ should be as well covered by the definition of standards. Some of them are already implemented: Statistical Data and Metadata eXchange (SDMX) is a full ISO technical specification.
To sum it up: if we consider that standards should apply to a given activity, it should be defined, beyond the GSBPM what is exactly included in this activity.
For instance, if we consider defining a standard on the data collection phase in Business statistics, it should first be defined more precisely what activity is precisely considered:
• is it the design? the development of the production system? the production phase (that is the data collection)?
• does it include the statistical specifications? the development of tools (and in particular questionnaires, IT tools, training of staff, monitoring of production).
• should we consider every production process in isolation, or with respect to the global production system? For instance, the NSI could have decided for organisational purposes, to monitor all data collection in one single division. For efficiency reasons, it could have considered to define common data collection modules for all types of surveys…
We have to emphasize that if we regard not only the statistical activity but the whole statistical business architecture, then the IT activity
• can be a part, a component of a statistical standard,
• can be part in the similar way to a financial standard or training standard but
• it might have the own standards for developing, supporting and managing its activity in the statistical organizations.
2 What kind of activity fields?
The precise and detailed content of a statistical standard can be different:
• it can refer to only one statistical domain, for example to the labour force survey, or
• it can refer all statistical domains like metadata guidelines or
• it can overlap a group of statistical domains like a handbook for business statistics.
Note: whereas the GSBPM is built on purpose independently of the domain for which statistics are produced, such a statistical domain may have influence on some features of the statistical process.
3 What kind of results?
ISO/IEC Guide 2: 2004[4] differentiates eight common types of standards based on purpose.
• A basic standard has a wide-ranging coverage or contains general provisions for one particular field, such as a standard for metal that can affect a wide range of products from cars to fasteners.
• Terminology standards are concerned with terms, usually accompanied by their definitions. The standards define words that permit industries or parties entering into a transaction to use a common, clearly understood language.
• Testing standards are concerned with test methods, sometimes supplemented with other provisions related to testing, such as sampling, use of statistical methods, or the sequence of tests. They are generally used to assess the performance or other characteristics of a product.
• Product standards specify requirements to be fulfilled by a product (or a group of products) to establish its fitness for purpose. Such standards can also address other issues, including packaging and labelling or processing requirements.
• Process standards specify requirements to be fulfilled by a process to establish its fitness for purpose. For example, a process standard could cover requirements for the effective functioning of an assembly line operation.
• Service standards, such as for servicing or repairing a car, establish requirements to be fulfilled by a service to establish its fitness for purpose.
• Interface standards, such as requirements for the point of connection between a telephone and a computer terminal, specify requirements concerned with the compatibility of products or systems at their points of interconnection
• Standards on data to be provided contain a list of characteristics for which values or other data are to be stated for specifying the product, process or service. This type of standard generally provides a list of data requirements for a product or service for which values need to be obtained.
It should be noted that these categories are not mutually exclusive. A process standard might contain the testing procedure and the description the product of the process as well.
All kind of these types have place in the statistical standards:
• Basic standard can be for example the code of practice, that gives direction to the statistical activities
• There are a lot of terminology standard for the common understanding the statistical concepts, terms for example statistical units in business statistics
• Testing standard can be in field of the questionnaire design, and developing application systems for statistical activities
• Product standards can be the different standard classifications and code lists, the description of standard registers like business register, the statistical outputs to be met the standard expectations. In the field of IT the products are for example the application systems.
• The different statistical methodologies (for sampling, estimating, etc.) can belong to the process standards. The standards for statistical workflow, the statistical processing steps (GSBPM) can be classified as process standards. In the field of IT support an example is the standards maintaining, supervising the IT system.
• Statistics belongs to the service branch. So the standards for different dissemination activities (standard for research rooms) can belong to this type.
• Interface standards can imagine for example to managing administrative sources, to pass statistical data to other systems (like SDMX).
• For standards on data to be provided typical examples are the metadata standards.
These classes can be joined into bigger groups:
• the description, regulation of certain activities (basic standards, testing standards, process standards, service standards) to the group of process standards
• the description, regulation of certain results of activities (terminology standards, product standards, interface standard, standards on data to be provided) to the group of product standards
4 Could IT tools be standards in statistics?
Application systems supporting the different statistical activities together with their documentation might become of standard use, if they were accepted to be standard by a recognized body and they were applicable for rather wide scope (for statistical domains, country or in other extension).
For example, packages like Demetra, BLAISE, and TAU-ARGUS could be considered as “products”. These products are not standards in themselves, but could be considered for some of them to be the result of an implicit pre-standard or of an implicit normative document.
This is the case for Demetra+, a seasonal adjustment tool that has been developed in order to take into account EU requirements in seasonal adjustment. The development, testing and maintenance has been conducted by many NSIs.
TAU-ARGUS, a package for statistical disclosure control, has been developed with EU funding in R&D projects, and has benefited from presentations and discussions at various meetings in Eurostat.
BLAISE is probably a tool that is more influenced by the CBS itself, although it is widely used.
Similar tools developed by one NSI or another could as well be exchanged and adapted to a wider ESS audience: CLAN, Bascula, and Calmar. Are there conditions to consider so that such tools could improve “standardisation”? Packages for assisted coding, etc…could be exchanged as well. “Specification languages” could as well be considered…The list could be extended: Eurostat developed[5] packages for data collection of the Intrastat statistics, for instance. This analysis will be developed in another document of this ESSnet paper.
This approach could be more systematic: the ESSnet CORE investigates, building on work done within ESSnet CORA, how to develop a generic approach to this.
There is as well some work done on the architecture of the Information systems. An attempt to define an agreed description of this architecture is on-going. One paper on the Generic Statistical Information Model (GSIM) is attached.
In Hungary within the Hungary Central Statistical Office (HCSO) we regard national standard
• our system for survey control - so called GÉSA system - that is meta data driven and independent from statistical domains. It manages and controls 132 data collections, which is the 98 % of data collections where the data supplier is an institution or individual agricultural farm.
• For data entry and validation our standard system, so called ADEL system that also meta data driven system and manages about 90 data collections integrated to the GÉSA system.
• For data collection via internet we use in this moment the so called KSHXML system, that also meta data driven system and manages 82 data collections integrated to the GÉSA and ADEL system.
5 The decomposition of statistical activities
If we regard the standards relating to the statistical activities, we could consider that the set of all activities could be decomposed along three dimensions:
• the first dimension could be the activity field of standard, the type of statistics to be produced (characterised by the statistical units - households, businesses, other, the frequency, the type of information - administrative or possessed by the statistical unit, the domain of investigation - social, economic, etc…)
• the second dimension could be the activity, the step - or the group of steps as described in the GSBPM
• the third dimension could be the professional aspect: purely statistical, methodological aspect, IT aspect (data model, functions, implementation, etc.), the organisational aspect (human resource, financial) , introduction, training aspect
The analysis is detailed in figure 2. The three parallelograms represent possible choices of activities to be standardised.
The green rectangle corresponds to the standardisation of the purely statistical aspects for the whole production process (all GSBPM steps) for one chosen statistical domain.
The pink parallelogram corresponds to the modelling aspect of steps 2 to 5 of the GSBPM for a set of statistics…
The yellow parallelogram corresponds to the activities in step 9 of the GSBPM, for all statistics and all statistical aspects (purely statistical, modelling, IT aspects).
This decomposition can be continued to take into account the different results of activities (different type of standards describing products and processes).
Figure 2: The decomposition of statistical activities
Coherence in standardisation
It is generally accepted that standards should be coherent. This implies that for a given activity or product, in a given context and a given year, accepted by a given “recognized body”, there should exist only one standard. Otherwise, two different standards would correspond to different degrees of order, and one of them would not be the optimum. If two different sets of rules, guidelines and characteristics lead to exactly the same level of order, then they should be included in the same standard and the choice between the two possibilities should be covered by the standard.
The optimum is that if for the same activity or product in a given context and for a given year the same standard is valid in the different level of official statistics. Or if the standards accepted by the different bodies for the same subject are different, then the standards have to be coherent with each other. There is the situation in the field of national account. (such as SNA, ESA and the Hungarian Method for National accounts).
Formally speaking, this means that the system of standards is defined by (e,B,E,F, date), where e is a Boolean variable that is equal to yes when a standard is exist, and no when no standard is available.
If we remember that E corresponds to an activity or a product, we can deal with coherence in a wider sense, i.e. if one activity E is made of two activities E1 and E2, does it mean that the standard corresponding to E will be the union of the standards corresponding to E1 and to E2 - provided they relate all to the same context? (If they do not relate all to the same context, these results will probably never hold).
In the above context, a standard that refers to a group of activities for a given statistical domain (for example Business statistics), and there are separate standards for some of the products and/or processes included into the standard (for example product standard for metadata or process standard for time series analyses), then the standard of the statistical domain must apply and refer in its document the standard of the existing standards.
In practice, the situation might be more complex. To check for consistency of standards, the use of a model like the entity-relationship model developed in WP1 might be necessary to keep track of existing standards and of the rules and guidelines included.
Classification according to the general business statistical architecture
Based on the previous considerations, a classification of statistical standards can be proposed according to the specific subject. The coding of groups tries to correspond to the ISO standard classification.
The classification of ISO standards (ICS) is a grouping of the standards by the field of activity.
Recoding of the statistical activities to a specific activity class will be required and we can propose the introduction of e.g. the code 04 for the statistical activity class.
The previous chapters dealt with the possible groups of the elements of statistical activities, the statistical business architecture, in that we can create standards.
In the point 3.1 we consider four strata of the business architecture. If we apply the ICS coding system, these can be the group of standards
• 04.010 the core of statistical activities aimed to deliver statistical products
• 04.020 the IT activity supporting statistics
• 04.030 the operating of statistical organisation
• 04.040 the statistical research and training activities
Figure 3: Standards of statistics
In the different groups can be different subgroups. As subgroups of the statistical activities, we propose to retain the classes of the GSBPM.
The standards can be general standards that overlap all statistical domains, or they can refer only one statistical domain or the group of statistical domains.
In the group of statistical standards we propose the next subgroups.
04.010 Standards of the statistical activity
04.010.10: Standards of specification of needs for statistics
04.010.20: Standards of design of statistical process
04.010.30: Standards for building the statistical process
04.010.40: Standards for collecting statistical data
04.010.50: Standards for processing statistical data
04.010.60: Standards for analysing statistical data
04.010.70: Standards for disseminating statistical data
04.010.80: Standards for archiving statistical data
04.010.90: Standards for evaluating statistical data
04.010.99: Other standards of the statistical activity
04.020: the IT activity supporting statistics
04.020.10: Standard application systems, software of statistics
04.020.20: Standards for IT developments in statistics
04.020.30: Standards for the operation and management of IT systems
04.020.99: Other standards of IT activity in statistics
04.030 Standards of the statistical organisation
04.030.01 Standards of the statistical organisation
04.040 Standards of the research and training
04.040.01 Standards of the research and training
The class “04.010 Standards of the statistical domain” contains all standards contains“ all statistical standard, that is related the main activity of a statistical organisation.
The group “04.010.10 Standards of the general statistical products“ includes all standards that describe the result of the statistical process (like statistical units, registers, measures, classifications, metadata, publications, statistical databases, product quality, etc.). These standards are not joined to a given statistical domain, fully or partially general for all statistical domains. The statistical product is regarded a wider context than the definition beyond, because it define product not only the statistical data, but metadata, classification, quality, statistical database as well.
The group of “04.020: the IT system of statistics“ sum up the IT standards supporting and realizing the statistical production
The subgroup of “04.020.10: Standard application systems, software of statistics” includes all standard IT system that are applied in the different steps of statistical processing flow. The extension of using the software can be different. The usage is standard, if the applications are independent from e.g. any subject matter domains or it can be use an important part of the subject matter domains. The IT systems can be either target application systems developing for statistical purposes to a given statistical procedure or we can use general commercial, software for carrying out a processing step.
This item includes standards on the products of IT developments, which are application systems, applications, and modules for the statistical business process model in order to conduct it
The subgroup of “04.020.20: Standards for IT development in statistics” include the standards those using is expected at the software or other IT development in the statistical system (project management, designing, programming, testing, documentation tools, etc.).
This item includes standards on the process of IT developments, which are application systems, application, and modules for the statistical business process model in order to conduct it.
The subgroup of “04.020.30: Standards for the operation and management of IT systems” includes the standards that are applied for working of the IT environment (like architecture, documenting, maintaining IT infrastructure, security policies, etc.).
This item includes the standards on the operation and management of IT systems e.g.: responsibility, security, issues, etc
The subgroup of “04.020.99: Other standards of IT activity in statistics” includes those standards that refer to the IT activity of statistics but can not be sorted to the previous subgroups.
The group 04.030 and the subgroup of “04.030.01: Standards of the statistical organisation, administration” includes the standards for operating a statistical organisation, implying the planning, financial activity, human resource, etc.
This item includes standards on statistical administration, which are support activities for the statistical business process e.g. planning of resources, financial activity, templates, etc.
The group 04.040 and the subgroup of “04.040.01: The standards of the statistical research and training” include the standards of the activity of statistical education and R&D.
This item includes the standards on the statistical research and training, which support creating new results in the statistical science and practice, and the better understanding and the knowledge of the above mentioned classification items.
Annex 1: Generic Statistical Business Process Model
Annex 2: Standards of statistics - topics and examples
Main subgroups of standards in statistics
04.010.10 Standards of general statistical products
04.010.20 Standards of general statistical methods (theoretical and practical methods)
04.010.30 Standards of general statistical production process
04.010.40 Standards of a given statistical domain
04.010.99 Other standards of the statistical activity
04.020.10 Standard application systems, software of statistics
04.020.20 Standards for IT developments in statistics
04.020.30 Standards for the operation and management of IT systems
04.020.99 Other standards of the IT activity in statistics
04.030.01 Standards of statistical administration
04.040.01 Standards of statistical training
Content of main topics (and some example)
04.010.10 Standards of general statistical products
|Description: This item includes the standards on “statistical products, which are generally, information |
|dissemination products that are published or otherwise made available for public use that describe, estimate, |
|forecast, or analyse the characteristics of groups, customarily without identifying the persons, organisations, or |
|individual data observations that comprise such groups”. (source: OECD Glossary of statistical terms) |
1) Terms of statistics
i. OECD Glossary of statistical terms,
ii. SDMX Metadata Common Vocabulary,
iii. SDMX Cross-domain concepts
2) Classifications, nomenclatures
i. SDMX Standard Code lists
ii. EU regulation on Classification of Economic Activities in the European Community (NACE) Rev 2
3) Statistical units, registers
4) Metadata descriptions for statistical data included description of methodology of statistics
i. 1/2009. (SK 1.) Statistical standard on the system of domains and on the description of domain’s methodology
5) Expectations on statistical publications
6) Expectations on accessibility, process ability of statistical data on internet, in research room, etc.
7) Product quality
i. VI/2008. (SK 2.) HCSO prescription for the calculation of product quality indicators
8) Expectations on saving, archiving statistical data
9) Prescriptions, regulations on interfaces to different systems
04.010.20 Standards of general statistical methods (theoretical and practical methods)
|Description: This item includes the standards on statistical methods, which are ‘structured approaches, to solve a |
|problem. Methodology is a set of research methods and techniques applied to a particular field of study. (source: |
|OECD Glossary of statistical terms) |
1) Survey design
i. I/2010. HCSO prescription for design of statistical survey
2) Questionnaire design
3) Methods and types of data collection
4) Methods of sampling
5) Response -non-response control
6) Measuring response burden
7) Methods and description of data editing
i. Methods of imputation
8) Methods of index calculation
9) Methods of weighting
10) Methods of statistical analysis
i. XXXII/2005. (SK 2.) HCSO prescription for the practice of seasonal adjustment
11) Disclosure control methods
12) Evaluation of result of data collection, non-response,
13) Process quality
i. 1/2010. Statistical standard on the quality directives for certain phases of statistical production flow in HCSO
14) Standards of documentation
i. XXI/2005. (SK 2.) HCSO prescription for the aspects of methodological documents
04.010.30 Standards of general statistical production process
|Description: This item includes the standards on statistical production processes ‘The processes are for |
|manipulating or classifying statistical data into various categories with the object of producing statistics. In |
|SDMX, "Statistical Processing" refers to a description of the data compilation and other statistical procedures to |
|deal with intermediate data and statistical outputs (e.g., data adjustments and transformation, and statistical |
|analysis)...’. (source: OECD Glossary of statistical terms) |
1) Phases and sub processes of statistical production process, GSBPM
2) Common standards for the different steps, blocks of statistical production process, e.g.:
i. Metadata for describing and driving the step,
ii. Information for quality indicators,
iii. Support statisticians in the step,
iv. Logging the step, etc.
3) Standard for given types of steps (data collection, data editing, etc.)
4) Relations among phases, sub processes, steps
5) Workflow standards
i. Standards for managing the phases, steps
ii. Standards for roles, logistic, organisation
04.010.40 Standards of a given statistical domain
e.g. Handbook on the design and implementation of business surveys
04.010.99 Other standards of the statistical activity
04.020.10 Standard application systems, software of statistics:
|Description: This item includes standards on the products of IT developments, which are application systems, |
|applications, modules for the statistical business process model in order to conduct it. |
1) Software developed for a given statistical block:
i. BLAISE,
ii. Argus,
iii. Demetra, etc.
2) Commercial, canned software:
i. SAS,
ii. SPSS,
iii. Interactive Reporting,
iv. Excel, etc.
04.020.20 Standards for IT developments in statistics
|Description: This item includes standards on the process of IT developments, which are application systems, |
|applications, modules for the statistical business process model in order to conduct it. |
1) Tools for project management, e.g.:
• PRINCE2 is a de facto standard developed and used extensively by the UK government and is widely recognised and used in the private sector, both in the UK and internationally. It embodies established and proven best practice in project management.
2) Tools for systems design and modelling, e.g.:
• The Unified Modelling Language™ - UML –
is OMG's most-used specification, and the way the world models not only application structure, behaviour, and architecture, but also business process and data structure. UML, along with the Meta Object Facility (MOF™), also provides a key foundation for OMG's Model-Driven Architecture®, which unifies every step of development and integration from business modelling, through architectural and application modelling, to development, deployment, maintenance, and evolution.
3) Software palette for the development of common application software (open source products, Oracle, SAS, Java, etc,) Environments (operation systems, browsers) in that the application system have to work.
4) Programming, naming conventions, data modelling standards,
5) Standards for transmission data metadata:
• Extensible Markup Language (XML) is a simple, very flexible text format derived from SGML (ISO 8879). Originally designed to meet the challenges of large-scale electronic publishing, XML is also playing an increasingly important role in the exchange of a wide variety of data on the Web and elsewhere. XML
• XBRL is a language for the electronic communication of business and financial data which is revolutionising business reporting around the world. It provides major benefits in the preparation, analysis and communication of business information. It offers cost savings, greater efficiency and improved accuracy and reliability to all those involved in supplying or using financial data. It is an open standard, free of licence fees
04.020.30 Standards for the operation and management of IT systems:
|Description: This item includes the standards on the operation and management of IT systems e.g.: responsibility, |
|security, issues, etc. |
1) Regulations for IT security, protection, documentation
• ITIL®(Information Technology Infrastructure Library) is the most widely accepted approach to IT service management in the world. ITIL provides a cohesive set of best practice, drawn from the public and private sectors internationally.
04.020.99 Other standards of the IT activity in statistics
04.030.01 Standards of statistical organisation, administration
|Description: This item includes standards on statistical administration, which are support activities for the |
|statistical business process e.g. planning of resources, financial activity, templates, etc. |
1) Oracle Financial Analyser
2) Book-keeping applications, etc
04.040.01 Standards of statistical research and training
|Description: This item includes the standards on the statistical training, which support the better understanding |
|and the knowledge of the above mentioned classification items. |
1) ESTP courses
2) Former TES courses, etc
Annex 3: Notes on the Generic Statistical Information Model (GSIM)
Summary
The recently released strategic vision document from HLG-BAS (High Level Group for Strategic Directions in Business Architecture in Statistics) discusses the role of GSIM and GSBPM (Generic Statistical Business Process Model) in supporting the “industrialisation” of statistics.
In simple terms GSIM can be envisaged as providing a basis for statistical organisations to agree on common terminology and definitions to aid their discussion on developing metadata systems and information management frameworks.
More specifically, it is anticipated GSIM will fulfil an essential role as a reference model that can be operationalised on a consistent basis when defining the information required to drive statistical production processes as well as when defining the outputs(e.g. statistical data) and outcomes (e.g. process metrics) from those processes.
In this way GSIM will support “designed in” (and, therefore, built in) interoperability between new methodological and new technical developments undertaken within the community/industry of producers of official statistics. This represents a much more efficient and effective approach than seeking to achieve interoperability only after discrete developments have been delivered.
Establishment of GSIM will therefore facilitate partnerships within the community to collaboratively develop new and improved capabilities, and then facilitate broader sharing of these new and improved capabilities across the community as a whole.
The next section of this document summarises the background to the GSIM collaboration, including
• the NSIs which are currently collaborating to produce initial drafts
• the expected relationship between GSIM and standards such as SDMX and DDI-L
• the context of GSIM as an initial major output which enables operationalisation of a common statistical information management framework
• other key collaborative forums and initiatives which are recognised as having a critical role in reviewing and shaping definition of GSIM
The document then explores current broad thinking about the structure of GSIM, including
• the Common Reference Model (CRM) Layer (which is analogous in nature to GSBPM), and
• the Semantic Reference Model (SRM) Layer (an additional level of formalisation and detail that is required in order to support consistent operationalisation)
It concludes with more information in regard to the proposed timeline for this work, including the planned points for engagement with the broader community of producers of official statistics to seek their input in co-ordinating and shaping the approach to GSIM.
Background
The idea of a GSBPM was discussed at the meeting of MSIS (Management of Statistical Information Systems) in April 2010. Two months later the inaugural meeting of the Informal CSTAT workgroup on stronger collaboration on Statistical Information Management Systems (CSTAT is the OECD Committee on Statistics) identified an essential role for GSIM in providing a consistent reference model when defining information required to drive statistical production processes, and output from these processes.
Six NSIs participated in this inaugural meeting
• Statistics Sweden (SCB)
• Statistics Norway (SSB)
• Statistics New Zealand (SNZ)
• Statistics Canada (StatCan)
• Office for National Statistics (ONS)
• Australian Bureau of Statistics (ABS)
Recommended ways forward identified by this workgroup were reviewed and endorsed by the Directors General of the participating agencies.
Operationalising GSIM was agreed to be the highest priority strategic enabler of efficient and effective collaboration in the development and sharing of statistical information management systems. This operationalisation was seen as progressing through associating GSIM with, for example, a commonly agreed to representation in XML. Rather than redundantly investing huge amounts of time and money developing such a representation “from a clean sheet of paper”, the approach would harness existing standards based representations wherever fit for purpose.
The workgroup identified SDMX and DDI-L (DDI-Lifecycle) as the key starting points in this regard. By relating GSIM to the information models associated with SDMX and DDI (neither of which covers the anticipated full scope of GSIM), a path to operationalising GSIM which draws extensively on SDMX-ML and DDI-L would be available.
The establishment and operationalisation process as a whole was termed OCMIMF (Operationalise a Common Metadata/Information Management Framework) by the workgroup. GSIM was defined as the initial major output which then enables operationalisation steps to be undertaken.
The OCMIMF “Opportunity Statement” as agreed by the inaugural workgroup meeting is attached to this document as Annex 1. More information about the collaborations, including OCMIMF, which were agreed at the inaugural workgroup meeting, is available via the MSIS wiki.
While participants in the workgroup committed to progressing work on OCMIMF “with pace and passion” they recognized this effort needs to be co-ordinated with other collaborative forums and initiatives and benefit from input from these.
Prominent examples include
• The METIS (Statistical Metadata) Group, jointly convened by UNECE, Eurostat and OECD and reporting ultimately to the Conference of European Statisticians
• Relevant ESSnets (Collaborative ESS - European Statistical System – Networks), such as
o CORA (Common Reference Architecture) and its successor CORE (Common Reference Environment)
o SDMX ESSnet, especially Work Package (WP) 2 related to MCV (Metadata Common Vocabulary) Ontology
The recently released strategic vision document from HLG-BAS (High Level Group for Strategic Directions in Business Architecture in Statistics) discusses the role of GSIM and GSBPM (Generic Statistical Business Process Model) in supporting the “industrialisation” of statistics. This vision is to be presented for discussion at the 59th Plenary Session of the Conference of European Statisticians (14-16 June 2011, Geneva).
Proposed broad structure of GSIM
There are often references to parallels between GSBPM and GSIM. The two are intended to work together, with the former providing a reference model for statistical business processes and the latter providing a reference model for information to be inputted to, used and produced by those processes.
The documentation of V4.0 (e.g. para 11) highlights that (at least at this time) GSBPM does not formally model attributes which are required in order to “operationalise” business processes in practice. This approach is entirely consistent with the (currently) agreed scope and intent of the GSBPM.
A key aim, however, is to consistently operationalise GSIM and to harness it to support “designed in” (rather than “attempted after the fact”) interoperability between new methodological and new technical developments undertaken within the community/industry of producers of official statistics.
This is seen as requiring GSIM to include an additional layer of detail compared with GSBPM. This would provide more formal “reference semantics”.
The figure on the following page illustrates this idea, where GSIM spans both the Reference Model Layer and the Information Model Layer. The “operationalisation” of GSIM (eg mapping to representations in SDMX and DDI-L) then supplies consistent connection between the Information Model Layer and the Physical Implementation Layer.
GSIM itself should remain distinct from, but connected to, recommended “operationalisation” of GSIM (eg representation, physical implementation). The recommended means of “operationalising” GSIM may change over time as technical standards and business practices commonly employed by producers of official statistics in their production processes evolve over time.
While GSIM itself may also need to evolve over time (probably following the same pattern as GSBPM in evolving more rapidly at first then achieving greater stability) it is important to its utility and uptake that it remains as stable as possible as a reference model. It should only change when statistical production processes of the future require, or produce, conceptually different/additional information – not when a next generation of processes require fundamentally the same information but in a different format (eg using a different technical standard).
Common Reference Model (CRM) Layer
[pic]
The Business Communication Diagram is intended as a simple (not technical) diagrammatic representation of GSIM. It is analogous to the much referred to diagram of GSBPM provided in Section IV of the documentation of GSBPM V4.0. For many business staff the diagram is their “day to day reference” in regard to GSBPM. Such staff members seldom (if ever) refer to the more detailed documentation. (It is debatable whether this is a good or bad thing – but it is a reality in any case.)
Similarly to GSBPM it is expected that GSIM will consist of multiple levels. The diagram for GSBPM, for example, shows
• Level 0 (statistical business process)
• Level 1 (nine phases of the statistical business process)
• Level 2 (sub-processes within each phase)
The documentation of GSBPM is seen as incorporating Level 3 in the form of descriptions of each sub-process.
The OCMIMF collaboration team is currently exploring different ways to describe and represent Level 1 and Level 2 of GSIM. Several possible representations have been identified by participants already.
For example, one approach has been based on briefly reviewing information models associated with SDMX, DDI-L, the MetaNet initiative, efforts by NSIs to define enterprise level information models etc and then identifying a consolidated superset of high level concepts/constructs from these
• another approach has been based on reviewing the text of the description of the GSBPM and identifying the information objects that it mentions as being associated with statistical business processes
Considering the top layer of GSIM from several perspectives (including checking it against the information objects that NSIs refer to in their actual systems/applications used to support the statistical production process) is seen as vital.
It is anticipated that during May 2011 early thoughts on the CRM Layer will be made available by the OCMIMF team (see below) and they will encourage review and input from all agencies and initiatives which have an interest in GSIM.
A key issue at the CRM Layer is terminology used to refer to, and describe, the information objects within the model. This requires a two level approach.
For example, if Agency A refers to “data elements”, Agency B refers to “variables” and Agency C refers to “data items” then
1. are the definitions used by the agencies consistent enough in meaning/concept to be represented by a single “box” in the common reference model layer?
2. if so, which label should be placed on the box?
Similarly to the success of the GSBPM, it needs to be recognised that this is a Common Reference Model. It does not require that each agency commits to “enforcing” each selected term as the preferred terminology for that agency’s internal purposes and that all existing internal documentation, repositories and user interfaces be updated accordingly.
The aim is to focus on agreeing the information objects (concepts) within the model and on a set of terminology that is acceptable for reference purposes – rather than ideal from every perspective. Such an approach is necessary and appropriate in order to achieve practical progress within a reasonable timeframe.
Semantic Reference Model (SRM) Layer
Defining and agreeing the SRM can progress most efficiently and effectively once there is an agreed working draft of the CRM Layer. The CRM sets out the high level information objects (including “common reference terms” to be used when referring to them) which will be defined in more formal, technical detail in the SRM. Attempting to progress development of the SRM without a foundation of common thinking in regard to high level objects and terminology would lead to a lot of talking at cross purposes.
That said, the refinement of the GSBPM took several years to complete. It is not necessary or appropriate to delay commencement of work on the SRM until the CRM has evolved to the same level of maturity as the GSBPM.
The intent is to commence work based on the initial agreed working draft of the CRM Layer. In effect, this first draft provides a basic (but essential) “common vocabulary” for the team in progressing work on the SRM. If/when there are subsequent changes to the CRM layer (eg changes to high level objects and/or to the terms used to refer to them) these should be readily factored into the SRM work.
It is recognised there are a range of existing well developed reference points for work on the SRM. These include
• information models associated with SDMX and DDI-L,
• MCV Ontology work undertaken within the SDMX ESSnet,
• outputs from the MetaNet initiative, and
• efforts by NSIs to define enterprise level information models
The agreed working draft of the CRM Layer will provide a framework for
• relating these existing reference points to the context of GSIM,
• harmonising across existing reference points where appropriate for the purpose of GSIM, and
• identifying gaps that need to be addressed in the context of GSIM
It is recognised that many of these reference points took several years to research, design, evaluate and refine. Given this wealth of existing intellectual investment and experience is available as a starting point, however, it is expected than an initial (potentially partial) draft version of the SRM will be available before the end of 2011.
Proposed Stakeholder Engagement
The draft stakeholder engagement timeline recognises the importance of sustained engagement with other collaborative forums and initiatives such as METIS and the CORE ESSnet. For example, starting from May 2011 it is currently planned there will be two rounds of engagement with these groups in order to “road test” ideas (at different levels of detail) related to the Common Reference Model before the team formally submits a draft of the CRM to the Directors General who commissioned the work.
Similarly active engagement is expected to then follow in regard to the Semantic Reference Model.
A more general approach to awareness raising and engagement will target forums such as the SDMX Global Conference, MSIS, IASSIST and ISI.
OCMIMF “Opportunity Statement”
The following opportunity statement was agreed at the inaugural meeting of the Informal CSTAT workgroup on stronger collaboration on Statistical Information Management Systems
|# |OPPORTUNITY |
| | |
|3. |OPERATIONALISE A COMMON METADATA/INFORMATION MANAGEMENT FRAMEWORK |
|MAJOR INITIATIVES |
|Major Initiative One |
|Develop a generic statistical information model (GSIM) |
|Develop requirements (what the model is supposed to achieve) |
|Investigate existing solutions and evaluate them against predefined criteria (incl. SDMX & DDI) |
|Build a draft model and test against real information/processes |
|Present the model to relevant formal and informal communities |
|Develop a plan for adoption and implementation |
|Major Initiative Two |
|Map SDMX and DDI to the GSIM framework |
|Major Initiative Three |
|Update and evolve the standards |
|Major Initiative Four |
|Implement SDMX and DDI in a test situation |
|Major Initiative Five |
|Operationalise the use of metadata as a driver for business processes (GSBPM) |
|- experimentation |
|STATISTICAL NETWORK BENEFITS |
|* Makes production process more effective * A small step for Statisticians, a giant step for Statistics * Enables |
|sharing * Key enabler for new information solutions |
|* Foundation for long term savings and operating costs * Increase user value |
2.1.1 Process for the Adoption and Implementation of Standards in the CEN Standardisation Process
1.0 Introduction
The European Committee for Standardisation (CEN) is an organisation which acts as a business facilitator in Europe, by removing trade barriers for European industry and consumers. It provides a platform for the development of European Standards and other technical specifications. The thirty one National Members work together to develop the voluntary European Standards (ENs). These standards have a unique status since they also are national standards in each of its thirty one member countries.
2.0 European Standard
An EN is a document that has been adopted by one of the three recognised European Standardisation Organisations: CEN, European Committee for Electrotechnical Standardisation or European Telecommunications Standards Institute. An EN is available in principle in the three official languages of CEN (English, French and German). A CEN standard represents a model specification, a technical solution against which a market can trade. It codifies best practice and is usually state of the art. CEN standards relate to products, services or systems. The standards are developed by consensus and reflect the economic and social interests of the CEN Member countries. The development of the standards is driven by business and each one is drafted by experts in the field. The CEN Standards are created by bringing together all interested parties including manufacturers, consumers and regulators of a particular material, product, process or service.
3.0 Standardisation Process
The following stages are involved in the formulation of a standard:
3.1 Proposal – Starting the work
An interested party is able to introduce a proposal for new work in the CEN. Most standardisation work is proposed through the National Standards Bodies.
3.2 Acceptance
The appropriate CEN Technical Committee makes a decision on the adoption of the proposal. An accepted standardisation project is allocated to one of its working groups for the drafting of a standard. One of the principles of standardisation in CEN is that once a project to develop a standard has been accepted, the National Standards Bodies put all national activity within the scope of the project on hold. This means they do not initiate new projects, nor revise existing standards at national level. This obligation is called 'standstill' and allows efforts to be focused on the development of ENs.
3.3 Drafting
The EN is then developed by experts appointed in a working group of the Technical Committee.
3.4 CEN Enquiry – public comment at national level
Once the draft of a European Standard has been prepared, it is released for public comment; a process known in CEN as the 'CEN Enquiry'.
3.5 Adoption by weighted vote
A final version is drafted which is then submitted to the thirty one CEN Members for a weighted Formal Vote. After its adoption, each of the National Standards Bodies publishes the new EN as an identical national standard and withdraws any national standards that conflict with it. The outcome of this is that one EN becomes a national standard in the other countries of CEN. This is unique in the world and with one common standard in all these countries and every conflicting national standard withdrawn, a product can reach a far wider market with much lower development and testing costs. This approach helps to build a European Internal Market for goods and services.
3.6 Review
A review is initiated by the Secretary of the Technical Committee after 4 years. It may be carried out earlier at the request of the Technical Committee, the Technical Bureau (BT), a CEN National Member, the European Commission, the European Free Trade Association Secretariat or the CEN Management Centre (CMC). If no Technical Committee exists, it is the responsibility of the BT to decide after close examination of the proposal prepared by the CMC.
Reference
•
Annex A – Diagram of the CEN Standardisation process
THE CEN STANDARDISATION PROCESS
[pic]
2.1.2 The ISO Standardisation Process
1. Background
1. The origins of the ISO occurred in 1946, when delegates from 25 countries met in London and decided to create a new international organisation “to facilitate the international coordination and unification of industrial standards.”
2. To date, ISO's work has resulted in over 16,000 International Standards, representing more than 620,000 pages in English and French (terminology is often provided in other languages as well).
2. Objectives
1. The main objectives for the ISO when applying standards are to make the development, manufacturing and supply of products and services more efficient, safer and cleaner and to facilitate trade between countries. The standards provide a technical base for health, safety and environmental legislation, and conformity assessment. They also allow people to share technological advances, good management practice and to disseminate innovation. The standards safeguard consumers of products and services and can make life simpler by providing solutions to common problems.
3. ISO Consensus
1. ISO standards are developed according to the principles of consensus where the views of all interested parties are taken into account. These could be manufacturers, vendors, users, consumer groups, testing laboratories, governments, engineering professions and research organisations.
4. The ISO Standardisation Phases
1. There are three main phases in the ISO standards development process and these are as follows:
• Expression of need - The first phase is when the need for a standard is expressed by an industry sector via a national member body. Once recognised and formally agreed, the first phase involves the definition of the technical scope of the standard and is usually carried out in working groups of technical experts from countries interested in the subject matter.
• Consensus and negotiation of detail - Once agreement has been reached on which technical aspects are to be covered in the standard, a second phase is entered during which countries negotiate the detailed specifications within the standard. This is the consensus-building phase.
• Formal Approval - The final phase comprises the formal approval of the draft International Standard. The acceptance criteria stipulate that approval by two-thirds of the ISO members that have participated actively in the standards development process, and approval by 75% of all members that vote must take place.
2. A number of stages are used in the development of the standards and these are included in the following table:
|ISO Stage |Detail |Equivalent ESSnet Standprep Proposal |
|Proposal |This is the first step in the development of an International |Adoption of standard |
| |Standard and it is the confirmation that a standard is needed. A | |
| |new work item proposal is submitted for vote by the members of the| |
| |relevant committees to determine the inclusion of the work item in| |
| |the programme of work. The proposal is accepted if a majority vote| |
| |in favour and if at least five members declare their commitment to| |
| |participate actively in the project. At this stage a project | |
| |leader responsible for the work item is normally appointed. | |
|Preparatory |This is usually when the working group of experts is set up for |Design and preparation |
| |the preparation of a working draft. Successive working drafts may | |
| |be considered until the working group is satisfied that it has | |
| |developed the best technical solution for the problem being | |
| |addressed. At this stage, the draft is forwarded to the working | |
| |group's parent committee for the consensus-building phase. | |
|Committee |As soon as a first committee draft is available, it is registered |Design and preparation |
| |by the ISO Central Secretariat. It is then distributed for comment| |
| |and voting, if required. Successive committee drafts may be | |
| |considered until consensus is reached on the technical content. | |
| |Once consensus has been obtained, the text is finalised for | |
| |submission as a draft International Standard. | |
|Enquiry |The draft International Standard is circulated to all ISO member |Design and preparation |
| |bodies by the ISO Central Secretariat for voting and comments | |
| |requested within a period of five months. | |
|Approval |The final draft International Standard is circulated to all ISO |Design and preparation |
| |member bodies by the ISO Central Secretariat for a final Yes/No |/ Communication |
| |vote within a period of two months. If technical comments are | |
| |received during this period, they are no longer considered at this| |
| |stage, but registered for consideration during a future revision | |
| |of the International Standard. | |
|Publication |Once a final draft International Standard has been approved, only |Communication/ Implementation |
| |minor editorial changes, if and where necessary, are introduced | |
| |into the final text. The final text is sent to the ISO Central | |
| |Secretariat which publishes the International Standard. | |
|Review of International|All International Standards are reviewed at the least three years |Maintenance and support Conformity |
|Standards – |after publication and every five years after the first review by |assessment |
|Confirmation Revision |all the ISO member bodies. |Archiving |
|Withdrawal. | | |
5. Conclusion
There are many similarities between the proposed standardisation process for the ESSnet Standprep project and that of the ISO standardisation process and some of the processes overlap.
Reference:
•
Annex A – Diagram of the ISO Standardisation Process
THE ISO STANDARDISATION PROCESS
2.1.3 Processes for the Adoption and Implementation of Standards: the Planned ONS Harmonisation Process
1.0 Introduction
The Context of UK Harmonisation
1.1 ONS independence
In July 2007, the Statistics and Registration Service Act received Royal Assent and brought in a new era of independence for both the Office for National Statistics (ONS) and for the broader UK statistical system. The Act enabled the creation of a new independent body, the UK Statistics Authority,(UKSA) which distances itself from ministers as a non-ministerial department, accountable to Parliament.
1.2 The United Kingdom Statistics Authority (UKSA)
The UKSA has oversight of the work of the Office for National Statistics (ONS) and is responsible for the independent scrutiny (monitoring and assessment) of all UK official statistics . It is required to monitor, and report on the production and publication of all official statistics (wherever produced), regardless of whether or not they are labelled as National Statistics.
1.3 Devolution and the Devolved Administrations
UK devolution created a national Parliament in Scotland, a national Assembly in Wales and a national Assembly in Northern Ireland. This process transferred varying levels of power from the UK Parliament to the UK's nations but kept authority over the devolved institutions in the UK Parliament itself. For Harmonisation this means not necessarily everyone using the same standard, but rather that the questions and standards that the user needs are used appropriately, in harmony. For example, the Scottish Government need to produce statistics that are right for Scotland and for ethnicity the different ethnic make up in Scotland must be reflected.
1.4 The Government Statistical Service (GSS)
The work of Harmonisation spans across the whole Government Statistical Service (GSS) which is a community of all those involved in producing official statistics in the UK. Led by the National Statistician it is a diverse network of people across many different organisations; including statisticians and those who may be part of other professional groups, part of a wider community of public servants providing high quality information, analysis and advice to decision-makers
1.5 Heads of Profession (HoPs)
The GSS Heads of Profession (HoPs) are responsible primarily for the statistical process and outputs within their specific Government Department. They develop and maintain the arrangements for pre-release access; advise the National Statistician on specific requests for access; and for ensure GSS policies are implemented. They have a strong role to play in Harmonisation by ratifying the proposals for harmonised standards.
2.0 The GSS Harmonisation Initiative
The harmonisation initiative aims to ensure that the same basic meaning for key concepts no matter when or how they are collected. It is important for a dataset to be comparable both within itself over time and with other datasets covering similar population characteristics. Harmonisation is used to establish common standard definitions and methods which are important in supporting comparability within the UK and internationally.
The standards are formulated by a Harmonisation Topic Group, which consists of a number of subject specific stakeholders (including representatives from the UK Nations) mostly for survey specific questions, and is used to provide input and comment during the planning, development and finalisation of harmonised standards.
When viewing harmonisation across the UK nations and internationally, the context of economic, social, legislative, institutional, cultural and other differences within the separate countries should always be considered.
3.0 Plans for highlighting potential topics requiring harmonisation
• To identify potential topics for harmonisation, workshops involving National Statistics Harmonisation Group (NSHG) members, the ONS Harmonisation team and wider interest groups are used to review the current package of standards and emerging issues.
• Tools such as SWOT analysis are used to clarify and prioritise harmonised standard requirements. These are used in the discussions around whether the new/revised harmonised standards work should be progressed further.
• The levels of data topic sensitivity for Devolved Administrations are highlighted, e.g., a red, amber, green system of indicators.
• A paper is then produced detailing the plans for the development of specific harmonised standards. This includes a detailed statement of user requirements and must be agreed by all stakeholders including the UK Nations. The paper includes a summary of stakeholder comments and gives a provisional view of key aspects including: consultation plan, development plan/timetable, potential issues such as data collection modes, etc. The paper is then presented to the harmonisation board for approval and agreements to proceed with the development of the standard. This approval is then ratified by a governing board of Heads of Profession.
3.1 Standards tested and further consultations with stakeholders
• Where applicable, proposed harmonised standards are field tested and issues raised are explored and fully detailed. Where this is not possible, a review by data collection methodologists is carried out.
• A summary report is produced to record the findings and issues highlighted from the testing phase along with the outcome of feedback received from stakeholders and UK Nations.
3.2 Endorsement requested from Harmonisation Board
• Harmonised standard proposals are presented to the harmonisation board for endorsement and a recommendation that the standard be ratified by the governing board
3.3 Ratification requested from the Governing Board
• Harmonised standard proposals are presented to the Governing board for ratification.
• A response paper is produced to clarify any issues and questions highlighted by GSS SPSC along with planned resolutions.
3.4 Presentation to Statistical Heads of Profession
• The Harmonised standard proposals are submitted to HoPs.
• Once HoPs have agreed the standard, it may be declared and published
3.5 Publication of ratified harmonised standard
• The standard is published on the National Statistics website
• A Harmonised question library is being developed to…
Annex A – The ONS Harmonisation Process Diagrams
[pic]
[pic]
2.1.4 DESTATIS Process (Paper outstanding)
Processes for the Adoption and Implementation of Standards: the key elements of the standardisation process for the ESS.
Chapter 2.2/2.3 COMBINED
Produced by Methodology, Standards and Register Department, GUS Methodology Department, and ONS Harmonisation.
1.0 Introduction
The aim of this paper is to present a standardisation process which could be adopted for use in the European Statistical System (ESS), within the framework of Workpackage 2 of the ESSnet StandPrep project.
A standard in itself is of no use unless it is implemented. A standard needs to be established, communicated, implemented, maintained and supported. Education and guidance on how to utilise a standard should be made available. When appropriate, the use of a standard has to be monitored and sanctioned. All of these management activities must be addressed in order to implement a standard.
The paper is based on the following sources:
• main findings from the document ‘Adoption of the Mandate of the Sponsorship on Standardisation’ taken from the 9th Meeting of the ESS Committee, Luxembourg on 18/19 May 2011;
• discussions based on presentations prepared by experts from CBS, DESTATIS, INSEE, and ONS on their own experiences in standardisation activities.
• a review of organisations covering International Standards.
2.0 Key Issues
2.1 Governance and Organisation
The first stage in the development of a standard is to establish an organisation of professionals with like minded aims, who do things in similar ways and who recognise the need to standardise the approach to their work. Such an organisation needs to be respected and internationally recognised (if the standard is to be used internationally) and should have the authority to ensure that the final proposal is universally adopted. The standardisation process should be managed in a formal, structured and open way, ensuring extensive opportunity for consultation and review with stakeholders.
An example of one such organisation is the International Standards Organisation (ISO), which is the world's largest developer and publisher of International Standards. The ISO is a network of the national standards institutes from 162 countries. There is one member per country, with a Central Secretariat in Geneva, Switzerland, that coordinates the system. It is a non-governmental organisation which enables a consensus to be reached on solutions that meet both the requirements of business and the broader needs of society. More information on ISO is included in Paper 2.1.1.
Another example is that of the European Committee for Standardisation (CEN) which is an organisation created to facilitate business in Europe by removing trade barriers for European industry and consumers. Its mission is to guide the European economy with issues linked to global trading, the welfare of European citizens and the environment. It provides a platform for the development of European Standards and other technical specifications. More information on CEN is included in Paper 2.1.2.
2.2 Procedures for the adoption of standards
Once the need for a standard is recognised and formally accepted, the next phase involves definition of the technical scope. This can be carried out in working groups of technical experts from organisations interested in the subject matter. The example from the ISO process shows that this step provides the confirmation that a standard is needed by a work item proposal being submitted for vote by the members of the relevant steering committees. The proposal is accepted if the majority of members vote in favour and if at least five members declare their commitment to participate actively in the project. At this stage, a project leader responsible for the work item is appointed and a working group of experts is set up for the preparation of a working draft. Successive working drafts may be considered until the working group is satisfied that it has developed the best technical solution to the problem being addressed. Consultation with stakeholders is crucial at this stage if the standard is to fulfil the widest set of requirements. Depending on the nature of the standard, there may be a need to carry out some testing or technical quality assurance of the proposals.
The ABC of Standardisation (Maureen A. Breitenberg, Standards Services Division Technology Services, National Institute of Standards and Technology, USA) highlights that when formulating standards, there are several considerations to be taken into account. The acceptance of a consensus is all important, ensuring that all views are heard and that the resulting standard is generally agreed by those involved. It also ensures the transparency of the process by recording the involvement of all parties, and providing advance public notice of a proposed standard.
The importance of balance must be taken into consideration, so that no one group’s interest dominates the approach and ‘due process’ must take place to ensure that anyone with a ‘direct and material interest’ has a right to express a position and to have that position considered or to appeal if necessary.
Openness and impartiality are also essential, as are effectiveness, relevance and coherence, in order to avoid the development of conflicting international standards.
The ISO has a number of members in developing countries with specific support needs and has developed technical assistance and training packages tailored to the variety of needs of its members and a training catalogue describing the courses it provides is available at iso/education_and_training.
Another way the ISO supports its members is by the ISO Action Plan for developing countries 2011-2015, funded by donors and other ISO members. The ISO view the technical assistance and training they provided as absolutely critical for keeping the ISO system primed and operating to the maximum efficiency.
CEN has introduced a helpdesk to provide on-demand advice and information to support European small and medium-sized enterprises on standardisation related issues. The helpdesk will introduce these enterprises to the information and business tools required to access the CEN system, and this will provide free information with links to brochures, presentations and access to experts via e-mail.
2.3 Repository for holding agreed standards
There needs to be an accessible repository to hold final agreed standards. The ISO provides a ‘Concept Database’ (ISO/CDB) () which makes the benefits of using standards easier to achieve. This database provides a platform for search, development and maintenance of concept content throughout the ISO standards portfolio. The ISO/CDB enables the search of concepts in three important categories, terms and definitions, graphical symbols and codes (country, currency, language and script). Another example is the Inventory of Global Statistical Standards, commissioned by the United Nations Economic and Social Council. The inventory is intended to provide a knowledge centre and reference system for international statistical standards. As a reference system, the inventory will also serve as an analytical tool for international organisations and groups, to help identify statistical areas requiring recommendations or standards as well as to identify overlapping standards. This inventory is being developed in phases and a report covering the project is available at:
2.4 Communication of standards
Once finalised, the standards need to be disseminated via publications, the internet, websites, etc.
At the ISO, once a final draft International Standard has been approved, only minor editorial changes, if and where necessary are introduced into the final text. The final text is sent to the ISO Central Secretariat which publishes it.
2.5 Support mechanisms for implementation of standards
There are a number of measures which can be undertaken to support the implementation of standards and these are:
• Training and education
• Technical assistance
• Helpdesks
• Mentoring
• Workshops
• Conferences and seminars
2.6 Conformity Assessment
Conformity assessment procedures may need to be adopted. The findings of a conformity assessment for a standard (e.g. identification of non-conformance and potential risks) will lead to the development of correction and improvement actions which should be carried out. Findings should also refer to decisions concerning whether a standard is up-to-date and all those found to be out-of-date should be archived.
Conformity assessment activities include:
• Sampling
• Testing
• Inspection
• Supplier's declaration of conformity
• Certification processes
• Management system assessments
• Registration processes
The UK Statistics Authority has a Monitoring and Assessment Team whose role is to systematically review statistical outputs against the Code of Practice. These assessments take place in accordance with a published programme, although ad-hoc assessments may be required from time to time in order to report on matters of public concern.
Self-evaluations are conducted using a standard template based on the Code of Practice and this forms the starting point for discussions between the Monitoring and Assessment Team and output managers. This information will be supplemented by any previously published reviews of the statistics and any information submitted by users or stakeholders.
The Authority's annual reports to the UK Parliament and devolved legislatures detail the outcomes of the assessments conducted in the previous year with their conclusions, as well as those planned for the coming year.
In addition, the ONS has developed a Quality Methods and Harmonisation Tool (QMHT) which has been designed as a 'stand alone' tool for managers to use to evaluate surveys and outputs. It offers a means of identifying areas needing further examination and asks the questions that a methodologist would. It provides a way of collating information on survey and output processes and methods to identify common practice. The tool uses many ideas from a previous Self-Assessment Checklist developed for the Government Statistical Service in the UK and also from a Eurostat model designed for surveys only.
3.0 Proposals for the stages to be incorporated into a process for the ESS.
In reviewing the approach of various ISOs, it has become apparent that there are a number of steps in the standardisation process which should be adopted. Once the need for a standard arises and the costs and benefits have been identified; the standardising organisation (in this case the ESS), should bring together stakeholders to communicate the decision and to canvass for contributors to develop the standard and establish consensus. The design and preparation stage along with the drafting of the standard then occur, and may involve testing the proposal or a technical quality assurance. Once finalised, the proposal is formally presented for public consultation and, if necessary, ratification by the governing body.
The adoption of the standard will involve publication and communication to highlight where the standard can be found. Implementation will involve support activities, such as training, and the standard will need to be regularly reviewed, revised, maintained and conformity assessed. If standards are superseded, then the previous versions must be archived.
The following is a list of proposed stages that should be incorporated into an ESS standardisation process.
|Proposals for the stages of an ESS Standardisation Process |
|1. |Need for a standard identified and |The first phase in developing a standard occurs when a requirement is identified and expressed |
| |defined |as a statement of user need. This stage includes identification of the aims of a standard and |
| | |its ability to satisfy recognised and future user needs, taking into account that it is much |
| | |easier to make a standard successful if users benefit directly and so the satisfaction of users |
| | |should be a decisive factor in taking decisions about the creation of standards. |
|2. |Costs and benefits of new standard |Alongside the statement of user need, there should also be a statement (where possible) of the |
| |identified |costs and benefits associated with the adoption of the standard. The related benefits and |
| | |projected costs should be detailed to illustrate the added value and potential burdens for both |
| | |data providers and users. Alternatively a ‘SWOT’ (strengths, weaknesses, opportunities, |
| | |threats) analysis could be carried out. A review of the process used to develop the standards |
| | |should occur, with the aim of ensuring the burden for all parties involved is minimised. |
|3. |Bring together stakeholders |The managing organisation should contact users, identify their needs and collect and collate the|
| | |feedback from users and stakeholders. Tools such as stakeholder analysis can be used to identify|
| | |the key stakeholders and users and the production of a consultation plan will help to highlight |
| | |all interested parties and their assigned status within the consultation process, the |
| | |consultation schedules and the framework for recording issues raised and related discussions. |
| | | |
| | |It is possible that the design or application of standards may differ in relation to different |
| | |modes of data collection or for different country conditions and these differences need to be |
| | |taken into account during the consultation stage. |
|4. |Design and preparation |This stage involves the preliminary investigations regarding technical specifications and legal |
| | |aspects related to the standards. Logistical management of standards and resources should also |
| | |be recognised. Preparation of draft documentation (including a description of procedures) and |
| | |creation of prototype standards occur at this stage. |
|5. |Acceptance by interested parties – |Preliminary acceptance of the proposal for standards by the managing organisation then takes |
| |consensus |place. With the ISO, the standard is circulated as a final draft to all members for a final |
| | |Yes/No vote within a stated time period. This is the formal presentation of the proposal and if|
| | |technical comments are received during this period, they are no longer considered at this stage,|
| | |but registered for consideration during a future revision of the standard. |
| | | |
| | |Once the standard has been approved within the standardising body, it is presented for public |
| | |consultation leading to the formal adoption and publication of the standard. |
|6. |Drafting the standard |Definition of the technical scope can be carried out in working groups of technical experts from|
| | |organisations interested in the subject matter. The example from the ISO process shows that this|
| | |step gives confirmation that a standard is needed via a work item proposal being submitted for |
| | |vote by the members of the relevant committees. The proposal is accepted if the majority of |
| | |members vote in favour and if at least five members declare their commitment to participate |
| | |actively in the project. At this stage, a project leader responsible for the work item is |
| | |appointed and a working group of experts is set up for the preparation of a working draft. |
| | |Successive working drafts may be considered until the working group is satisfied that it has |
| | |developed the best technical solution for the problem being addressed. Consultation with |
| | |stakeholders is crucial at this stage if the standard is to fulfil the widest set of |
| | |requirements. Depending on the nature of the standard, there may be a need to carry out some |
| | |testing or technical quality assurance of the proposals. |
|7. |Testing the proposal of a new |Where applicable, proposed standards should be tested and any issues raised should be explored |
| |standard |and fully detailed, with a summary report produced to record the findings and highlight any |
| | |issues from the testing phase along with the outcome of feedback received from stakeholders. |
|8. |Technical QA |It may be necessary to have the proposed standard quality assured by a relevant external body to|
| | |assess its completeness and fitness for purpose. |
|9. |Formal presentation of the proposal |The formal proposal for the standard should detail all the main stages of development and |
| | |include all questions and issues raised during the process along with the planned resolutions. |
| | |The results from the testing of standards and feedback from consultations with stakeholders |
| | |would also be included. |
| | | |
| | |The document should include a proposed timetable for finalising the standard and should outline |
| | |any key milestones along with risks, issues, dependencies and contingencies. |
|10. |Ratification |If the standards body reports to a higher authority, it may be necessary to have the agreement |
| | |to the new standard ratified by this higher body. |
|11. |Public dissemination |A communication plan should be drawn up to ensure that information about the adoption of the new|
| | |standard is disseminated as widely as possible, especially to stakeholders, and any remaining |
| | |issues identified and dealt with. |
|12. |Adoption of the standard |Consensus on common rules and understanding of standards expressed by members of working groups |
| | |or other bodies involved in the using of standards should then take place. Recognition for |
| | |coherence of the given standards and links with other strategic issues that have an impact on |
| | |standards are clarified. The mandate for the responsible authority (steering committee) for |
| | |governance of standards and for establishing its duties is established. |
|13. |Publication/Communication |Preparation of the dissemination of standards via publications, internet, websites, etc. At the |
| | |ISO, once a final draft International Standard has been approved, only minor editorial changes, |
| | |if and where necessary are introduced into the final text. The final text is sent to the ISO |
| | |Central Secretariat which is responsible for publication. |
|14. |Placing in a repository |There needs to be an accessible repository to hold the final agreed standard. |
| | | |
| | |The ISO provides a ‘Concept Database’ (ISO/CDB) () which makes the benefits of|
| | |using standards easier to achieve. This database provides a platform for search, development and|
| | |maintenance of concept content throughout the ISO standards portfolio and ISO/CDB enables search|
| | |of concepts in three important categories, terms and definitions, graphical symbols and codes |
| | |(country, currency, language and script). |
|15. |Implementation |This sees the application of standards within the domains for which they were created, e.g. |
| | |surveys, classifications, statistical regulations. |
|16. |Support |There are a number of measures which can be undertaken in support of implementing the standards.|
| | |These are: |
| | | |
| | |Guidelines |
| | |Training/Education |
| | |Technical assistance |
| | |Helpdesks |
| | |Mentoring |
| | |Workshops |
| | |Conferences and seminars |
|17. |Review and improvement |Periodic review of the standard should take place at regular intervals. Improvement actions |
| | |refer to amendments in old standards and may also help to identify the need for new standards |
| | |when the old standards are not able to be improved further. |
|18. |Revision |Revision of a standard may become necessary due to advances in technology etc. At the ISO, a |
| | |number of the International Standards developed by ISO technical committees require updating or |
| | |revision. To do this, a competent body which has the requisite infrastructure for ensuring the |
| | |effective use of these international agreements is appointed. These bodies are designated by |
| | |ISO to serve as maintenance agencies or registration authorities. Details concerning the review|
| | |stages used by the ISO are available at: |
| | |
| | |table.htm |
|19. |Maintenance |Management of standards in the repository is an important issue and this includes the loading of|
| | |standards into the repository, checks of completeness against the description of the standards, |
| | |and updating and developments based on improvement actions formulated during the conformity |
| | |assessments. Also maintaining motivation incentives for the further development of standards |
| | |should be considered. |
|20. |Conformity assessment |Conformity assessment in comparison to the specification of standards may be necessary, for |
| | |example, assessment of quality criteria (quality components of ESS, costs and effectiveness, |
| | |other universal quality criteria taken from the intrinsic nature of standards). |
| | | |
| | |Findings are recorded to highlight the results from conformity assessments of standards (e.g. |
| | |identification of non-conformance and potential risks) and the formulation of correction and |
| | |improvement actions which should be carried out during the maintenance of standards. Findings |
| | |should also refer to decisions concerning whether a standard is up-to-date and all those found |
| | |to be out-of-date should be archived. |
| | | |
| | |Conformity assessment can include sampling and testing as well as the inspection or the |
| | |supplier’s declaration of conformity. There are also certification processes along with |
| | |management system assessments and registration processes that can be applied. |
|21. |Archiving |Provision must be made for the storage of standards which are out-of-date with no further |
| | |application. |
4.0 Main Findings
• There are a number of logical and sequential steps in the standardisation process used by internationally recognised standards bodies which could be adopted for use in the standardisation of statistical processes in the ESS. Statistical standards should be managed by a comprehensive standardisation process. The process begins with the identification of the need for a standard and finishes with conformity assessment and archiving of standards, when they are no longer applied.
• Effective consultation with stakeholders is crucial if the standard is to fulfil the widest set of requirements.
• Depending on the nature of the standard, there may be a need to carry out some testing or technical quality assurance of the proposals.
• Standardisation activities should provide maximum functionality with the minimal number of tools.
• Interdisciplinary teams that consist of subject experts including methodology, quality, and IT specialists should be involved in the standardisation activities and in the standardisation process.
• The importance of balance must be taken into consideration; in that no one group’s interest should dominate the process.
• There are a number of measures which should be undertaken in support of developing standards such as technical assistance and training packages.
• It is possible that the design or application of standards may differ in relation to different modes of data collection or for different country conditions and these differences need to be taken into account during the consultation stage.
Annexes
• Annex A compares several different approaches to the standardisation process.
References
• ISO Organisation.
• CEN Organisation.
• The ABC of Standardisation (Maureen A. Breitenberg, Standards Services Division, Technology Services, National Institute of Standards and Technology USA).
• The Six Dimensions of Standards: Contribution towards a theory of standardisation – E. Baskin*, K. Krechmer* and M. H. Sherif - Communications Standards Review Palo Alto, CA, USA and AT&T Labs, Paris, France.
• Adoption of the Mandate of the Sponsorship on Standardisation for 9th Meeting of the European Statistical System Committee, Luxembourg 18/19 May 2011; Annex: 10.
Annex A – Comparison of Standardisation Processes
|PROCESS STAGE |CEN |ISO |DESTATIS |ONS |PROPOSED
ESSNET
PROCESS | |1 |NEED FOR A STANDARD ARISES |X |X | |X |X | |2 |COSTS AND BENEFITS OF NEW STANDARD IDENTIFIED |- |- | |X |X | |3 |BRING TOGETHER STAKEHOLDERS |X |X | |X |X | |4 |DESIGN AND PREPARATION |X |X | |X |X | |5 |ACCEPTANCE BY INTERESTED PARTIES – CONSENSUS |X |X | |X |X | |6 |DRAFTING THE STANDARD |X |X | |X |X | |7 |TESTING PROPOSAL OF THE NEW STANDARD |X |X | |X |X | |8 |TECHNICAL QA |X |X | |- |X | |9 |FORMAL PRESENTATION OF THE PROPOSAL |X |X | |X |X | |10 |RATIFICATION |X |X | |X |X | |11 |PUBLIC DISSEMINATION |X |X | |- |X | |12 |ADOPTION OF THE STANDARD.
|X |X | |X |X | |13 |PUBLICATION /COMMUNICATION |X |X | |X |X | |14 |PLACING IN A REPOSITORY |X |X | |X |X | |15 |IMPLEMENTATION |X |X | |X |X | |16 |SUPPORT |X |X | |- |X | |17 |REVIEW AND IMPROVEMENT |X |X | |X |X | |18 |REVISION |X |X | |X |X | |19 |MAINTENANCE |X |X | |- |X | |20 |CONFORMITY ASSESSMENT |X |X | |- |X | |21 |ARCHIVING. |X |X | |X |X | |
3.1 Development of a common IT Package in the ESS - The Demetra Experience
1.0 Introduction
Demetra is one of the very few packages that have been successfully developed in common in the ESS. The underlying work started in 1996 and the first version of Demetra was issued in 1998. It then evolved and has just been replaced by Demetra+. The intention was not for Demetra to become a common package for the statistical production in the ESS. Raoul Depoutot was the Eurostat coordinator of the initiative between 1996 and 1998, and has reported on the dynamic of the development and the difficulties faced.
2.0 The objective of the initiative
NSIs were using different methods for seasonal adjustment (SA) that were not considered satisfactory for the use of seasonally adjusted series that was developing around the start of the EMU. A number of NSIs were beginning to produce seasonally adjusted series after the introduction of the STS regulation and the development of quarterly national accounts. NCBs were conducting their own seasonal adjustment, and had traditionally gained more experience in this domain (usual in economic analysis). It was therefore necessary to compare methods used by NSIs to find out which one should be used as a common method by all NSIs.
3.0 Problems raised
Eurostat hired a first specialist in seasonal adjustment, who had compared methods (BV4, Dainties, X11-Arima, TRAMO-SEATS) used by MS and Eurostat on a subset of ESS data series. A report was produced which ranked the competing methods and showed that two methods were clearly not adapted for EU needs (BV4, used in one single MS, and Dainties used in the Commission). TRAMO-SEATS was considered to perform better on average than X11- ARIMA. Eurostat discontinued the use of Dainties, BV4 as it was considered to be below acceptable quality for the majority of MS series. Member States did not approve the result of the comparison between TRAMO-SEATS and X11-ARIMA.
The theoretical approach was felt to be very demanding. The underlying academic basis required high skills in mathematics and statistics (stationary and non stationary processes, filtering, etc). Only Eurostat and one or two NSIs had recruited statisticians with a satisfactory qualification.
Academic specialists were associated to the process, in order to provide assistance in the choice of methods. However the situation was quite complex as different methods were based on different hypotheses for modelling the time-series and the unobserved components (trend, cycle, seasonal and noise) raised the difficult problem of testing non-nested hypotheses. It appeared there could be data series for which method A was performing better than method B, whereas for other data series it would be the contrary. The debate by academics was based on published papers, using a small set of data series from very few countries. The debate required input from highly skilled statisticians but they were not available in the short-term in NSIs.
The Eurostat team proposed to have a pragmatic approach and to develop a business case based on the following:
• the objective was to focus on data series produced by NSIs;
• recommended methods would have to be accessible to statisticians with technical skills not higher than those typically available in NSIs;
• to define a standard process for seasonal adjustment that would be applied to all data series;
• to consider under which circumstances more skilled seasonal adjustment should be performed.
The objective for which the business case was defined consisted in defining a method that provided a sufficient quality for all series at a reasonable. This was different from the objective of academic specialists who aimed at the best possible solution for each series.
It was then decided to perform comparisons of available SA methods in each participating country. Due to practical and partly technical problems, many NSIs were not in a situation to use both methods. In particular, one of them (TRAMO-SEATS) was then only available as a Fortran program, without any user interface. Furthermore, the list of parameters was quite complex in both methods, and some countries that could perform the comparisons did not use parameters that were compatible with both methods (this referred to the need to test in a fair way).
4.0 Overcoming the difficulties
In order to overcome difficulties, it was decided to develop a package that was designed for use in applied research. This package would provide a user friendly interface for both methods and assist in the comparison of results of both methods. The need to produce fair comparisons was also taken into account. It was also clear that Eurostat did not intend to cover the costs for maintenance of a package that would be intended for production of statistics.
This phase highlighted some interesting lessons that could be of interest for future work in standardisation.
The development started with a survey of the current situation in NSIs concerning seasonal adjustment. It was observed that the description was very unclear when there was no centralisation (or at least internal coordination) at the NSI level. Only very partial information could be collected. It was suggested that emphasis should be placed on the interfaces, i.e. the possibility to read data series from existing databases and to store the results in these databases. An important finding was that the information received about the current situation mainly concerned the methodological units, and that details covering the production units were not sufficiently reported.
We had to face a situation in which the methodologists did not represent the concrete situation of production in statistical departments. There was far too much interest in sophisticated analysis, which was not compatible with the resources available for the seasonal adjustment of data series in general. The interest for the cost of such work methods was not shared by the majority of delegates, who focused on high quality SA. Hence, the hypotheses for the comparison of methods were in practice relatively biased. It should be noted however that SA methods are based on models that use a great number of parameters. In practice, modifying all these parameters can have an influence on the results. It was decided to provide some default values for all these parameters in Demetra.
5.0 The benefits of delivering a common package
Initially an important benefit was that it improved mutual understanding. Those who could not use TRAMO-SEATS before the interface, had now got easier access and could run their first simulation.
Providing visual presentation of results was also quite helpful, whereas seasonality is a complicated concept from a mathematical point of view so its visual representation is quite easy. In the end it is a matter of “smoothness” and “regularity” of a graph.
The use of default parameters helped a lot, as using such defaults, users could perform the first applications and compare similar approaches with both packages. In a sense it was closer to the operational results than with the initial packages, which required considerable skill.
Statisticians could easily replicate with national series what was shown by the team of experts at Eurostat, as they had hired a second high level expert with a PhD in seasonal adjustment methods and an expert focusing on aspects linked to the use of packages.
The dogmatic approach that was developing when the discussion was focused on theory and scientific methods was quite limited when the discussion could mention results from simulations and concrete achievements in practice. It became obvious that an important proportion of statisticians relied on default values for parameters in everyday practice and this helped to highlight the importance methods being user-friendly.
6.0 Towards standardisation in seasonal adjustment?
It should not be considered that the aim of Demetra was to provide a standard tool. It was rather a tool to assist in the negotiation for the adoption of common good practices in seasonal adjustment. It might have developed later into a tool fit for such standardisation.
However, the version of Demetra available at the end of the 1990s cannot be considered as a standard for seasonal adjustment. The package made available the two SA methods that were used by most NSIs in the world. So it did not really incorporate rules or guidelines that had an influence on statisticians. The main outcome was probably to transform TRAMO-SEATS (an experimental tool developed outside NSIs) into a package that now has an equal footing with X12. The package was developed by the US Bureau of the Census who are/were the historical leaders in SA.
In parallel, Eurostat and the working group on seasonal adjustment developed guidelines for seasonal adjustment. At the beginning of 2008, they were updated and adopted at a very high level by the ESS. Some standards now appear to have been adopted, but it did require more than 12 years. The current rules make it possible to use either TRAMO-SEATS or X12 ARIMA. Such a standard does not reduce a lot the degrees of freedom available to statisticians.
Author: Raoul Depoutot (ESSnet STAND PREP)
3.2 ARGUS as a De Facto Standard for Disclosure Control
Introduction
The software package ARGUS consists of two parts. Both parts deal with disclosure control. The package μ-ARGUS is concerned with disclosure control of microdata, whilst τ-ARGUS is used for the protection of tabular data. This paper is a short description about how ARGUS became a de facto standard for disclosure control.
2. History
ARGUS was initiated in 1993 by Statistics Netherlands. The first promising results led to a close cooperation between Statistics Netherlands, Eurostat and other countries. With funding from Eurostat, The Netherlands, UK and Italy produced a first version of both packages within the ‘4th Framework SDC-project 1998’.
The next development was the production of further developed packages within the ‘5th Framework CASC-project’ with the participation of Germany, Spain, UK and The Netherlands. Further development was supported by Eurostat via Centres and Networks of Excellence (CENEX) and ESSnet projects.
Present situation
New developments are now under the control of Statistics Netherlands. New versions are distributed via the CASC-website and are free for users. New funding will lead to more progress. Each package contains one user interface for the different underlying programs. These programs can be developed separately and are able to communicate with each other with the help of text files.
4.0 Lessons learned
The software architecture should facilitate distributed developments. The responsibility for development and maintenance should not depend on one person or one institute. There is a need for a permanent governance board and continuous funding.
Conclusion
ARGUS is an example of a local initiative which was recognised by the international community and developed with external funding and international participation. The present governance and funding is not well arranged.
To stimulate successes like ARGUS it is needed to:
• Stimulate local initiatives
• Recognise promising initiatives
• Develop an international co-operation
• Implement good governance and funding
3.3 Using XBRL in a Statistical Context – the case of the Dutch Taxonomy Project
Statistical offices are responsible for the publication of important national economic indicators such as gross national product, national accounts, structural business statistics and short-term indicators. For producing those indicators statistical offices rely on data provided, either directly or indirectly by companies and institutions (UN Economic and Social Council 2004). The received data reflect the economic activities of those companies.
Relevant data about the economic activities of companies are traditionally collected by statistical offices through paper or electronic forms in which companies provide the requested data. Statistics Netherlands is also increasingly relying on tax office data, such as value added tax and profit tax data.
In the past various attempts have been made by Statistics Netherlands and other statistical offices to supply companies and their software providers with a digital data exchange format. Examples of such attempts are the International Trade Statistic data format (Centraal Bureau voor de Statistiek 2007), The Teler initiative (Knuppel and Kunzler 2001), the EDIsent tool (Piebinga 1999), the Dutch "Elektronische Heerendiensten" initiative (EIM 2001) and the EGS-POS pilot (Ministry of Economic Affairs 2005). The electronic data exchange format can be used by software providers to export the data required by the statistical office out of various administrative systems. Experience has shown that the implementation rates of these initiatives are low. This is probably caused by the fact that software providers do not expect a profitable return on investment.
The business case for implementing those standards is made unattractive by two factors. First of all, the content and layout of the digital formats is often explicitly described in documents called message implementation guides. The software provider has to interpret those guides and implement the requirements into the software. Changes in either content or layout mean that those guides and software have to be updated. It takes precious IT-development time to correctly interpret, implement and test the digital exchange format and updates.
Another critical factor in the lack of acceptance of digital exchange formats imposed by statistical offices, is that they are not the only organisations (in the eXtensible Business Reporting Language (XBRL) community called "regulators") issuing such formats. Customs, tax offices, banks and insurance companies are just a few examples of regulators issuing their own, often unique data exchange formats. Since a statistical office by itself does not always provide enough potential customers (and customer pressure), the urge to implement the statistical digital exchange format is low for software providers.
Crucially, regulators often require similar or partly overlapping sets of data. So despite a multitude of different layouts for the existing data exchange formats, often the same administrative facts are required. Standardisation of the digital exchange format of the regulators could provide the leverage needed to make the business case for the software providers profitable. Secondly, harmonisation of concepts between regulators is a challenge to reduce response burden.
XBRL may very well be this standard format. Developed and maintained by a consortium of regulators, accountants and software builders, it can offer a link between the data kept in book-keeping systems and the data terms of regulators, such as national statistical and tax offices. In the Netherlands, XBRL has been introduced and used for this purpose. The initiative behind the use of XBRL is called the Dutch Taxonomy Project. It is a joint effort of the Dutch Ministry of Finance and the Ministry of Justice (2003). It aims at reducing the administrative burden on companies in the Netherlands to a large extent. Statistics Netherlands, the Dutch Tax Office and the Dutch Chambers of Commerce are the regulators involved in the Dutch Taxonomy Project. It is likely that the project will achieve its goals by harmonising legislation surrounding mandatory regulatory reporting and by introducing a joint supported single electronic data format for highlighting the data requirements of regulators.
3.4 SDMX (Paper outstanding)
CORE ESSnet: Principal Standardisation Efforts
The principal aim of the Common Reference Environment (CORE) project is the detailed design and prototype implementation of an architecture supporting the execution of statistical business processes. Such processes are defined in terms of services calling each other according to principles of modern information systems design. Hence, a first important notion of CORE is the definition of statistical processes in terms of identified services. This definition step is supposed to be performed by statisticians and indeed the services to be selected are statistical services abstracting from the specific IT implementations.
This is a relevant standardisation step for National Statistical Institutes (NSIs). The statistical user of CORE defines the process in terms of available statistical services. Moreover, CORE explicitly supports the classification of such services according to the Generic Statistical Business Process Model (GSBPM) standard. This is delivered by enriching the services available in the environment with a GSBPM tag. As in CORE, a statistical service is in principle mapped to several IT services; the possibility of tagging through GSBPM will enable searches to be carried out to retrieve information, for example, “all the IT services implementing the 5.4 Impute subprocess of GSBPM proposal”.
A further significant standardisation effort of CORE concerns data exchanges. One of the aims of CORE is to standardise the data flow underlying a statistical process in terms of both: (i) a common information model; and (ii) a unique technological transport format.
The CORE information model has been released by the CORE ESSnet as a dedicated deliverable (‘2.2 Generic statistical information model’). It is important to note that the CORE information model is not intended to represent business objects (enterprises, households, etc.), but it is instead focused on characterising the level of aggregation of data exchanged between services of statistical processes. In relation to this, ‘statistics’, ‘population’, ‘unit’ are possible metadata specified by the CORE information model. It should be noted that not being focused on business object representation, the CORE information model is complementary to the Generic Statistical Information Model (GSIM), and not actually overlapping. More specifically for CORE, the details of the particular information which is being communicated between services in a specific case is not relevant. The CORE information model operates in a sense as a ‘postal envelope’ and is used when passing information between services rather focusing in detail on the specifics of the information being communicated (i.e. on what is inside the envelope). The CORE information model explicitly takes into account ‘Business concept recognition’ as a requirement and therefore GSIM data will be recognised as such when exchanged between services.
From a technological perspective, CORE data are represented by means of XML technologies and are therefore further supporting standardisation. Moreover, CORE ESSnet (and previously Common Reference Architecture (CORA)) considers “technology independence” as a basic principle. This choice is indeed motivated by the consideration that each National Statistical Institute (NSI) has its own technologies and no approach imposing one unique set of technologies would have been successful. Instead, CORE aims to be able to “wrap” existing NSI services in whatever technology they are written, so acting more on the side of standardisation at the design level (e.g. exchanged data, clear communication interfaces, etc.) than at the implementation level (i.e., by imposing a specific technology).
-----------------------
[1]
[2] TR= technical report, Amd.= amendment, Cor.= correction)
[3] Source: Bo Sundgren: Developing and implementing statistical metadata systems,
[4] ISO/IEC Guide 2: 2004, Standardisation and related activities -- General vocabulary, provides general terms and definitions concerning standardisation and related activities.
[5] : I do not know if these packages have been adapted to the evolving IT environment
-----------------------
Product standards
Product standards
Process standards
Product standards
Process standards
Product standards
Standards for research and training of statistics and supporting activities
Standards for the operating stratum of a statistical organisation
Standards for
the stratum of IT support
Standards for the stratum of
the statistical domain
Professional
aspect
Statistical domains
9
8
7
6
5
4
3
2
1
GSBPM
Proposal
Acceptance of Proposal
Adoption by Weighted Vote
Agreement requested from Governing Board
Drafting
of Proposal
CEN Enquiry
Reviewed
after
4 Years
Publication - National Standard all Countries
Enquiry
Committee
Proposal
defgqwyz‘’åȪŒ?tgWG9/9hâ?aJmH sH jhâ?U[pic]aJmH sH hohM&l5?CJaJmH sH hoh,.5?CJaJmH sH hˆLn5?CJaJmH sH ho5?CJaJmH sH h¸wÏh¸wÏmH sH ;hoh,.6?B*[pic]OJQJ\?^JaJmH nH pPreparatory
Publication
Approval
Review
UK Nations
• User Requirements
• Consultation Plan
• Development Plan
• Timetable
• Key Issues
• Costs / Benefits
Topic Groups
Endorsement requested from Harmonisation Board
Harmonisation Plans
Gov Depts
Harmonisation Proposals
Process standards
Process standards
Further development of standards
Other Stakeholders
Harmonisation Process
AGREED STANDARDS PUBLISHED
Plans for new/revised harmonised standards agreed by GSS SPSC
Development of standards taken forward within agreed timetable
in consultation
Endorsement requested from Harmonisation Group
Ratification requested from Governing Board
Agreement requested from Governing Board
Standards tested and further consultation with stakeholders
Proposals for harmonised standards finalised
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.