77 institutional level and/or at the process or product level. On the process or product level there will be an option to: (a) apply all relevant TDQAF principles to all processes or products; (b) apply selected principles to all processes or products. During implementation, NSOs will conduct a data user-producer dialogue to provide information on quality and use its findings and conclusions alongside results from quality assessments and audits. Through this meeting, all members of the National Statistical System will commit to continually assessing, improving and reporting on the quality of official statistics, as well as on the quality of data and statistics used in the production of official statistics as required. Tanzania Data Quality Assurance Framework will be implemented at the NSOs and throughout the entire NSS. The TDQAF will also be applied to all data and statistics produced outside of the NSS that are disseminated as deemed appropriate and required, extend to data and statistics that are disseminated jointly with other statistics producers that are not members of the NSS with the help and support of a member of the NSS or that are used for government decision-making, as deemed appropriate and required. NSOs will also develop or review the existing subject matter quality assurance frameworks accordingly. The selection quality management across NSS will be based on the Generic Statistical Business Process Model (GSBPM) as recommended by UN NQAF manuals and NSOs will support members of NSS on their use. GSBPM has been used for the description and quality assessment of process based on surveys, census, administrative records and other or mixed sources. (See Annex I).
4.6 Assessment and Reporting The primary objective of quality assessment and reporting is to establish mechanisms designed to prevent, mitigate, and evaluate issues that may occur during the statistical process, which could affect the quality of statistical outputs. Implementing robust quality assessment approaches strengthens the organization's reputation as a professional and credible producer of high-quality data. Methods and tools for statistical quality assessment will comprise quality indicators (for both products and processes), quality reports, user satisfaction surveys; and external and self-assessments while labelling and certification will be looked upon as advanced practices. The use of these methods in an efficient and cost-effective manner requires that they be used in combination with each other. NSOs will use quality reports as the basis for
78
audits and user feedback. Assessment and reporting of data quality will be done by
considering the following components:
4.6.1 Quality Indicators
National Statistics Offices will identify quality indicators in order to measure compliance
with the respective quality principles and requirements. The QSTWG will define and
develop quality indicators to measure compliance with the respective quality principles
and requirements. The specific and measurable elements of statistical practice will be used
to characterize the quality of statistics. During the development of quality indicators,
those linked to GSBPM, IMF DQAF, UN NQAF among others will be reviewed and used
where applicable.
4.6.2 Quality Reports The quality reports will typically examine and describe the outcome of quality assessment according to components or dimensions (quality principles) that NSOs have used to define their products’ fitness for purpose. The reports will convey the necessary information to enable users to assess the quality of the product. Different user groups will be clearly identified and presented with different subsets of quality indicators. In the optimal case, quality reports will be based on quality indicators and presented according to a standard reporting structure to facilitate comparability.
4.6.3 Obtaining Feedback from Users User feedback is a crucial element in the set of information needed for a comprehensive quality assessment. NSOs will regularly consult with its users about their needs and perceptions of quality, take them into account in the quality assessment exercise and follow up with the users, for example through meetings (e.g., focus group discussions) or in a more formal way by using user satisfaction surveys.
4.6.4 Conducting Assessments Systematic self-assessments, external evaluations, internal and external quality audits, and peer reviews will be the primary tools used to conduct quality assessments within the NSS. The QSTWG will develop self-assessment checklists to facilitate a systematic evaluation of the quality of the statistical production process. The self-assessment will be
79
carried out by members within their respective functional areas at both the NSOs and
NSS, with support from the QSTWG.
Regardless of the assessment context, the results will be used to measure the degree of
compliance with established requirements and to identify both strengths and weaknesses.
This information will then be leveraged to enhance the quality of statistics produced both
within and outside the NSS. Moreover, this approach will demonstrate transparency
regarding the extent to which data quality standards are being achieved across NSS.
4.6.5 Labeling Labelling indicates the extent to which a set of quality standards are adhered. The attachment of a label requires a procedure to guarantee that the information is appropriate and true. This is for designating statistics as official or assessing their adequacy in terms of quality reporting. Thereafter, predetermined labels will accompany the various levels of fulfillment of quality standards with corresponding explanatory notes for each label. The methods and procedures for labelling will be developed by NSOs.
4.7
Assuring Continuous Quality Improvement
By implementing a quality approach following the different processes described above,
NSOs will define a framework for continuous quality improvement. If the new
information on quality that becomes available will always be considered in the statistical
outputs and statistical production processes. A cycle of continuous improvement of the
quality of the statistics produced will be established as an integral part of the statistical
agency's working practices.
80 ANNEX I: GENERIC STATISTICS BUSINESS PROCESS MODEL (GSBPM)
81
ANNEX II: PROCESSES DURING DATA QUALITY ASSESSMENT
The successfully of TDQAF need to have a proper working modality whereby the Quality
Unit will follow during the data quality assessment. The assessment will be done within
45 days by following these steps.
i)
Decision on the product to be assessed: Statistics products to be assessed should
be identified for every year;
ii)
Plan of the data quality assessment: Quality unit should develop a plan for
conducting assessment
iii)
Receive notification on the products to be assessed and readiness for the
assessment: the assessed department should receive notification on the products
that will be assessed from the quality unit;
iv)
Roadmap for assessment exercise: The quality unit will issue roadmap for
assessed department on each product which is planned to be assessed;
v)
Undertake orientation before assessment: The QATWG will get an orientation
on data quality assessment in order to know and understand the items that will
be needed during the assessment;
vi)
Undertake a pre-assessment: This will be undertaken before the actual
assessment in order to identify if the required items are available;
vii)
Conduct the actual assessment: The quality assessment of the product will be
undertaken by QATWG in order to be assured that the produced product follow
2024 Data Quality Assurance Framework;
viii) Making a follow up: After finishing actual assessment, the QATWG will follow
all unfinished requirements to the responsible unit, division or department;
ix)
Compilation and report writing: all necessary requirements will be compiled
and the quality assessment report will be written;
x)
Validation of the report by data quality management: This will involve quality
unit and key stakeholders;
xi)
Finalization of the report: finalise report by considering observations, comments
or suggests which were proposed;
xii)
Reporting according to structure: submission of the report will follow the
reporting structure of the NBS/OCGS;
xiii) Publishing the final report to the NBS/OCGS website.
82 The following diagram summarizes the work flow through which data quality assessment process will follow:
83 References
- United Nations Statistics Quality Assurance Framework (UN-SQAF), UN Statistics Quality Assurance Framework Including a Generic Statistical Quality Assurance Framework for a UN Agency New York, September 2016
- African Union, (2015), A Draft Statistical Quality Assurance Framework for the African Statistics System
- Generic Statistical Business Process Model GSBPM (Version 5.1, January 2019)
- Kenya National Bureau of Statistics (2022), Kenya Statistical Quality Assurance Framework
- National Bureau of Statistics (2014), Data Quality Assessment Framework for Tanzania Mainland
- National Institute of Statistics of Rwanda, (2014), National Quality Assurance Framework
- Nizhni N., (2014), High-level Workshop on Modernization of Official Statistics, Russian Federation
- Office of the Chief Government Statistician (2014), Zanzibar National Quality Assessment Framework
- Statistics Botswana, (2020) Gaborone Botswana Data Quality Assessment Framework, Private Bag 0024
- UN NQAF Self-Assessment Checklist, 2019
- United Nations Children’s Fund, (2021), Data Quality Framework, UNICEF, New York
- United Nations, (2019), National Quality Assurance Frameworks Manual For Official Statistics
- Statistics Quality Assurance Framework, United Nations Conference on Trade and Development (UNCTAD), Geneva, May,2019
- A Draft Statistical Quality Assurance Framework for the African Statistics System, Statistics Division Economic Affairs Department, African Union, May 2015
- Quality Assurance Framework of the European Statistical System, Version 2.0, 2019
- The National Quality Assurance Framework for Public Statistics (NQAF /PS), NIGER, National Statistics Council,
- IMF’s Data Quality Assessment Framework (DQAF)
- United State Census Bureau, (2003) Tool for Assessing Statistical Capacity
- https://atlan.com/data-quality-framework/
- https://unstats.un.org/unsd/methodology/dataquality/tools/
- The Statistics Act, CAP 351
- The Statistics Act No. 9 of 2007 and its associated Statistics Regulations of 2018 (Zanzibar)
- System of National Accounts 2008
- Tanzania Statistical Master Plan Phase Two (TSMP II) 2022/23-2026/27
84