66 3.4.5 Principle 18: Assuring coherence and comparability Principle 18 highlights the importance of consistency in the development, production, and dissemination of statistics. NSOs and Statistical Units must ensure that related data, originating from various sources, can be seamlessly combined and utilized jointly. Moreover, statistics should maintain comparability over time and across different geographical areas. This principle is realized through three quality requirements as elaborated in Table 3.18.
67 Table 3.18 Principle 18: Assuring coherence and comparability Requirement
Element to be assured
Description
18.1
International,
regional
and national standards
are used with regard to
definitions,
units,
variables
and
classifications.
Statistical agencies promote the adoption of national, regional or
international statistical standards.
Adoption and adaptation to these guidelines
need to be done by the statistics agency. The use
of
statistical
classifications
(ISIC,
COICOP,
COFOG, ISCO, CPC, HS) by NSOs serves as an
evidence
Guidelines,
a
common
repository
of
statistical
concepts,
definitions of units and variables and classifications and other
mechanisms exist.
The compliance with international, regional or national standards
for statistical production are periodically assessed. Any deviations
from these standards are made identified and included in the
publicly available metadata, along with reasons for such
deviations.
18.2
Procedures or guidelines
are in place to ensure
and
monitor
internal,
intra-sectoral and cross-
sectoral coherence and
consistency.
Statistics derived from different sources or with different
periodicities (e.g. monthly, quarterly and yearly) are compared
and any differences are explained and reconciled, as appropriate.
Active SWGs, meeting regularly with NSS
stakeholders is among the key actions towards
addressing such issues. The presence of schedule
of meetings for SWGs, and stakeholders serves as
a reference
Cooperation and the exchange of knowledge between individual
statistical programs and domains is promoted.
Process specific procedures and guidelines are available to ensure
that outputs are internally coherent.
Before new statistics or statistical programs are being launched
the conceptual and methodological relationship with existing
statistics is analysed.
Statistical outputs are compared with results of other statistical or
administrative sources that provide the same or similar
information on the same subject matter, and divergences are
identified and explained to users.
Internal procedures or guidelines are developed in order to
ensure and monitor internal coherence and consistency.
Procedures and guidelines are developed in order to ensure that
results from different sources can be combined. Compliance is
periodically assessed.
68 Requirement
Element to be assured
Description
18.3
Statistics
are
kept
comparable
over
a
reasonable
period
of
time
and
between
geographical areas.
Changes in methods of data compilation are clearly identified,
described and analysed to facilitate the interpretation of the
results.
Reports of statistical activities need to be kept
consistent over time. The agency needs to have a
specific report format for each type of statistical
product. When revisions are made trends should
be reproduced
Quality reporting includes a section on the assessment of internal consistency and comparability over time and with related statistics. Breaks in the series are explained and the methods for ensuring reconciliation over a period of time are made publicly available. Effects of changes in methodologies on final estimates are assessed and appropriate information is provided to users. Significant changes in the society and phenomena to be measured are reflected by appropriate changes to concepts, classifications, definitions and target populations. Differences within geographical areas or at the country level due to different concepts or methodologies are explained.
69 3.4.6 Principle 19: Managing metadata NSOs and Statistical Units should provide information covering the underlying concepts and definitions of the data collected and statistics produced, the variables and classifications used, the methodology of data collection and processing, and indications of the quality of the statistical information in general, sufficient information to enable the user to understand all of the attributes of the statistics, including their limitations. This Principle is managed by three quality requirements as elaborated in Table 3.19
70
Table 3.19
Principle 19: Managing metadata
Requirement
Element to be assured
Description
19.1
The metadata management system of
the statistical agency is well defined
and documented.
A strategy, guidelines and procedures are in place for
metadata management and dissemination.
The agency needs to have guideline for
definitions of all indicators and key concepts.
A compendium of statistics terms, concepts
and definitions serves this purpose
Metadata management is recognized as responsibility of
all staff.
19.2
Metadata are documented, archived
and
disseminated
according
to
internationally accepted standards.
International, regional, national or internal standards are
used for metadata documentation, management and
archiving.
Some
internationally
recommended
platforms
for
microdata
and
metadata
dissemination such as NADA can be
adopted. Tanzania adopted this platform for
the management and dissemination of micro
and metadata
Procedures are in place to ensure that metadata are
documented according to standardized metadata systems,
and regularly updated.
Metadata are made available at the same time as the data
and statistics to which they pertain.
The dissemination of metadata is tailored to different
needs, such as those of producers and users of statistics.
A systematic way for archiving metadata is available
which also ensures that they are accessible for reuse in the
future.
A glossary of statistical concepts is publicly available.
19.3
Staff
training
and
development
programs are in place on metadata
management and related information
and documentation systems.
Process managers are trained to properly document the
data and describe the relevant processes.
The training plan should also address the
issue
of
metadata
management
and
documentation.
Statistical agency staff participate in international metadata forums
71 CHAPTER FOUR IMPLEMENTATION OF QUALITY ASSURANCE FRAMEWORK 4.0 Quality Management The management of quality is an integral part of the management of every program in the NSOs and an important component of management as a whole. Both statistics offices have the mandate to play the role as a co-coordinating agency, within the NSS to ensure that quality official statistics are produced. The NSOs and other stakeholders of NSS prepare the guidelines to ensure the production of quality statistics in the country. These NSOs are responsible for spearheading the development of the national quality assurance framework in consultation with the stakeholders in the national data ecosystem. The TDQAF will guide the production of statistics as well as outline the criteria for monitoring and assessing the quality of statistics produced in the country. It will be an assurance that the players in the data ecosystem commit to continually assess and improve the quality of statistics produced in the United Republic of Tanzania. NSOs are responsible to ensure that the framework is adopted by national legislation and provision of support in its implementation across the national data ecosystem. It is anticipated that the Framework will be used in a different way by both providers and users of official statistics. For the producers of statistics, especially managers and staff responsible for different statistical processes, the framework provides a means of assessing the quality of different processes and products they are responsible for and identifying where improvements and changes are needed.
4.1 Institutional Arrangement NSOs will set up specific institutional arrangements and roles for data quality management. This will involve identification of responsible staffs for quality as well their clear roles and responsibilities around quality management. The institutional arrangement will involve four levels of hierarchy. The lowest level starts with Internal Quality Management Team (IQMT) in each NSOs. The IQMT will operate under the supervision of the managers responsible for quality statistics issues who, in turn, work under the supervision of the directors responsible for quality statistics in both offices. The Quality Statistics Technical Working Group (QSTWG) comprises quality focal points from NSS will be established in each NSOs and chaired by
72 the Manager or Directors responsible for quality. The NSS quality advisory board of the United Republic of Tanzania comprising members from NSS will be established and chaired by either head of Statistics Offices from NBS and co-chaired by OCGS and vice versa. A pictorial representation of this institutional arrangement is shown in the following diagram.
QUALITY COORDINATION STRUCTURE
73
4.2
Quality Unit
The quality unit is responsible for quality management or coordination within the NSOs in
order to maintain sufficient capacity to lead and support the implementation of quality
management initiatives throughout the NBS/OCGS and to support other NSS members
and producers of official statistics if required. As per current organization structures, the
quality unit in NBS is the Research and Standard Department and in OCGS is the Division
of Standard and Coordination of Statistics and Research. Other major statistics producers
within the NSS will be encouraged to establish their own quality unit to support the work
on quality within their organizations.
4.2.1 Internal Quality Management Team In order to assure the quality of official statistics within NBS/OCGS, The Internal Quality Management Team (IQMT) comprised members from quality unit and NBS/OCGS departments will be established. The members of IQMT will serve as quality champions or focal points in their respective department and support TDQAF implementation throughout the NBS/OCGS. The IQMT will serve as a forum in which quality related issues in the various aspects of NBS/OCGS operations can be addressed at both the strategic management and the operational level. The IQMT will also serve as a mobilization mechanism for quality management initiatives, such as documentation workshops or specialized training workshops for improving quality, among other things.
Joint meeting between NBS and OCGS will be conducted to discuss on the quality issues for Tanzania. The meeting will be chaired by directors responsible for quality from NBS/OCGS. The issues that will be raised will be presented to Statistician General and Chief Government Statistician.
4.2.2 Quality Statistics Technical Working Group In order to assure the quality of official statistics within NSS, the Quality Statistics Technical Working Group (QSTWG) will be established. The QSTWG comprised members from NSOs (internal quality management team comprises representative from each department) and respective representatives from NSS. The QSTWG will serve as: • The quality champions or focal points in their respective department and support TDQAF implementation throughout the NSS;
74 • A mobilization focal point for quality management initiatives, such as conducting training workshops for improving quality in their respective sectors.
The QSTWG will meet on quarterly basis to discuss data quality issues for NSS. The meeting will be chaired by Managers or Directors responsible for quality from NBS/OCGS.
75 4.2.3 Quality Assurance Managers To mainstream quality, the NSOs have designated the managers responsible for quality management who will be responsible for: - a) Establishing the comprehensive quality assurance plan defining all the quality- related activities, quality assessment and quality indicators to be implemented in NSOs activities; b) Prepare a budget for undertaking the planned assessments; c) Mobilise resources for quality assessment; d) Coordinate the QSTWG and IQMT; e) Review quality assessment reports and submit to the director responsible for quality management; f) Prepare training awareness on data quality assurance or management.
This intend to embed quality assurance directly into the core operations of each functional area under the NSOs by empowering the quality assurance managers to ensure quality standards and practices are upheld and consistent throughout the NSS. Other major producers of official statistics will be encouraged to establish quality managers or quality focal points.
4.2.4 Quality Assurance Directors The directors will be from the department which is overall in charge and management of issues pertaining to data quality management in NBS/OCGS. Usually, these directors are concerned with statistical standards and methods, among other issues. Under the aspect of quality management, the directors will be responsible for: a) Ensuring that data quality assessments are conducted timely and reports produced accordingly; b) Mobilizing resources and/or allocating budget required for undertaking planned data quality assessments during a specified time period; c) Advising the Heads of NSOs regarding to the findings of the quality assessment report and recommending means of further improving the quality of statistics.
76 4.2.5 NSS Quality Governance Board National Statistics Offices will establish a Quality Governance Board to ensure data quality throughout the entire NSS. Members of this group will compose directors from NSOs and respective representative directors from NSS. A primary task of this board is to ensure the statistical outputs meet the quality standards in all aspects of statistics development, production and dissemination. In this function, it may raise quality concerns, assess existing statistics or formulate requirements for additional statistics and analysis. Formulation of this board will regularly identify strengths and weaknesses; and continue to advise for more improvement.
4.3 Training of Staff National Statistics Offices will undertake adequate training to the quality management team and representatives from each directorate in NSOs, respective representatives from NSS, Quality Assurance Managers, Quality Assurance Directors and the Quality Governance Board. The training will focus on providing knowledge on basic concepts, objectives and tools of data quality assurance and the assessment framework that show a clear understanding of the quality principles, requirements and elements that are applicable to statistical outputs and processes.
4.4 External and Internal Communication National Statistics Offices will explicitly communicate its commitment to high quality and continuous improvement to its stakeholders in the form of a declaration on quality. The declaration will state the principles that guide the approach of NSOs to manage data quality, standards and commitments in the production of official statistics. The declaration on quality will be officially launched, be visible on the NBS/OCGS website and actively introduced and promoted internally and externally to all stakeholders through quality seminars, campaigns etc. An important instrument of communication such as external quality reports, which will address the needs of various user groups will be used.
4.5 Development of Implementation Strategy and Action National Statistics Offices will develop a short-term action plan for quality improvements covering the next data production cycle of the financial year and parallel to a mid and long-term strategy and action plan. Generally, quality assurance will be applied at the