en-1749018255-Tanzania Data Quality Assurance Framework (TDQAF) for Official Statistics_June2024.pdf

Type: Document | Status: ready

55 Requirement Element to be assured Description
to society’s emerging information needs. Cooperation with the scientific community and owners or holders of new data sources is established to experiment with and pioneer the use of these data sources.
serves the purpose of existence of strategy for addressing the utilization of new data sources

Possibilities of exploiting new data sources are regularly discussed by management. 14.4 User satisfaction is regularly measured and systematically followed up. User satisfaction surveys and user studies are regularly carried out and analysed. There should be a schedule for conducting a User Satisfaction Survey regularly. The user satisfaction survey report serves as the evidence Improvement actions arising from the user satisfaction surveys and user studies are identified and implemented. User satisfaction surveys include questions on the opinions of users about metadata availability. Measures to assess satisfaction of main users with particular products are in place (e.g. specific user satisfaction surveys and indicators on product level).

56 3.4.2 Principle 15: Assuring Accuracy and Reliability NSOs and Statistical Units should develop, produce and disseminate statistics that accurately and reliably portray reality. The accuracy of statistical information reflects the degree to which the information correctly describes the phenomena it was designed to measure, namely, the degree of closeness of estimates to true values. This principle is achieved through three quality requirements as elaborated in Table 3.15.

57 Table 3.15 Principle 15: Assuring Accuracy and Reliability Requirement Element to be assured Description
15.1 Source data, integrated data, intermediate results and statistical outputs are regularly assessed and validated. Systems for assessing and validating source data, integrated data, intermediate results and statistical outputs are developed and managed. Benchmarking, revisions, and rebasing need to be conducted whenever new data are available. Computations should rely on acceptable standards and methods. Benchmark and revision reports serve as evidence Data are systematically checked and compared with data from other sources and over time. Results of statistics are compared with other existing information in order to ensure validity. 15.2 Sampling errors are measured, evaluated and documented. Non- sampling errors are described and, when possible, estimated. Procedures and guidelines are available on how to measure and manage (e.g. to reduce or balance) errors.
Each sampling survey program report should be accompanied with the technical report which explains errors committed during the program implementation Sources of possible sampling errors are identified and described. Sampling errors are measured and evaluated. Non-sampling errors (errors from all sources, such as response errors, coverage errors, errors linked to measurements, processing, analyses etc.) are identified, described and evaluated. Errors are analysed to identify improvement measures. Information about the sampling and non-sampling errors is made available to users as part of the metadata. 15.3 Studies and analyses of revisions are carried out and used to improve data sources, statistical processes and outputs. Preliminary and revised data and statistics are clearly identified. Special new studies and further analysis of existing data need to be done when revisions, rebasing and benchmarking exercises are to be conducted. Trade margins, service surveys, NGOs survey are examples of this kind of studies

Explanations about the timing, reasons for and nature of revisions are made available. The revision policy follows standard and transparent procedures. Information on the size and direction of revisions for key indicators is used to improve the statistical processes. Information on the size and direction of revisions for

58 Requirement Element to be assured Description
key indicators is provided and made public.

59 3.4.3 Principle 16: Assuring timeliness and punctuality Principle 16 emphasizes the importance of timely and punctual dissemination of statistics by NSOs and Statistical Units. Timeliness pertains to the speed at which data and statistics are released to users following the reference date or period's conclusion, while punctuality concerns adherence to promised or advertised release dates. This principle is achieved through four quality requirements as explained in Table 3.16.

60 Table 3.16 Principle 16: Assuring timeliness and punctuality Requirement Element to be assured Description
16.1 Timeliness of the statistical agency’s statistics comply with international standards or other relevant timeliness targets. The timeliness of the statistical agency’s statistics complies with dissemination standards of international organizations such as the International Monetary Fund (IMF) or other relevant timeliness targets (e.g. requirements for Agenda 2030). International, regional and other dissemination standards (SDDS, GDDS) should be adhered for statistics production. Examples of publications which comply with these standards are quarterly and annual GDP and monthly CPI Divergences from international timeliness targets are monitored and actions are taken to comply with these if targets are not met. The overall trade-offs between timeliness and other dimensions of quality (e.g. accuracy, cost and respondent burden) are given consideration when setting targets. 16.2 The relationship with data providers is managed with respect to timeliness and punctuality needs. There are agreements with data providers on the planned delivery dates and delivery format. Statistics release calendar could be used as a tool for communicating with data providers on when data are disseminated Procedures are in place to ensure the effective and timely flow of data from providers and to statistical agencies. Follow-up procedures are in place to ensure timely receipt of data from providers. 16.3 Preliminary results can be released when their accuracy and reliability is acceptable. The possibility and necessity of releasing preliminary data for key statistics is evaluated, while also considering data accuracy and reliability. Preliminary results need to be shared with key stakeholders for their validation before final release. Minutes confirm the undertaking of stakeholders' validation serve as evidence
When preliminary statistics are released, they are clearly identified as such. Users are provided with appropriate information on the quality of the preliminary statistics. Preliminary results are revised according to the established revision policy. Final results are clearly distinguished from

61 Requirement Element to be assured Description
preliminary results. 16.4 Punctuality is measured and monitored according to planned release dates, such as those set in a release calendar. Punctuality or the rate of punctuality (i.e. rate of statistics published on time) is measured according to what is set in the release calendar at least 3 months ahead of publishing the relevant statistics. Statistics release calendar, schedule of work program and implementation reports can be used as a benchmark for assessing punctuality
Information on the punctuality of the released statistics is discussed by management and made available to users.

62 3.4.4 Principle 17: Assuring accessibility and clarity Principle 17 underscores the imperative for NSOs and Statistical Units to ensure transparent and accessible dissemination of statistics, facilitating ease of discovery and retrieval. Statistics should be presented clearly, promoting understanding among users, and be equally accessible in diverse formats adhering to open data standards. Additionally, access to microdata for research, while upholding statistical confidentiality, should be facilitated through established policies. This principle is achieved through seven quality requirements as elaborated in Table 3.17.

63 Table 3.17 Principle 17: Assuring accessibility and clarity Requirement Element to be assured Description
17.1 Statistics are presented in a form that facilitates proper interpretation and meaningful comparisons. Statistics are presented in a clear and understandable manner. Simple language and presentation should be used to disseminate statistics. Whenever available, trends should also be presented. Key finding reports are among the examples that facilitate easy interpretation of statistics Guidelines that describe the appropriate content and preferred formats and style (layout and clarity of text, tables, and charts) of the agency’s outputs are available to authors of statistical publications and databases. Published statistics are open for free use and re-dissemination, given that reference is made to the responsible agency.
Staff training and development programs are in place on writing about statistics (for press releases, publication highlights or other explanatory texts). Up-to-date methodological documents (on concepts, scope, classifications, basis of recording, data sources, compilation methods and statistical techniques), as well as quality reports and the work program of the statistical agency are made available to the public. Explanatory texts that accompany the statistics are reviewed for clarity and readability. Meaningful comparisons are included in the publications when appropriate. Preliminary and revised data are identified and explained in published statistics. Metadata needed to understand and use the statistics are published together with the statistics.
A policy for archiving published statistics is in place. 17.2 A data dissemination strategy and policy exist and is made public. The public are made aware that custom-designed outputs, statistics not routinely disseminated, and longer time series can be provided on request when feasible, and they are instructed how the data can be ordered. These outputs are made public if possible. A set of the following documents can serve for purpose: o Data dissemination and access policy

64 Requirement Element to be assured Description
Catalogues of publications and other services are made available to users. o Communication strategy o Release Calendar All these need to be posted on the website While official statistics are normally free and accessible for everyone, statistics that need to be produced on request might have a cost corresponding to the extra work it requires. Pricing of special requests is fully transparent. A strategy has been developed and agreed upon with stakeholders for the release of anonymized data and microdata. 17.3 Modern information and communication technology is used for facilitating easy access to statistics Statistics are disseminated in various ways suitable for all different users, with the agency’s website providing a central entry point. All statistical reports and data should be posted to a website, portals, dashboards following proper standards of disclosure and anonymization. The use of mobile device app to disseminate statistics also serve for purpose Users are able to extract data from statistical databases through public interfaces in the most appropriate and common formats (xlsx, csv, html, etc.). Statistical data can be accessed via an Application Programming Interface (API).
Statistics are disseminated in ways that facilitate re-dissemination by the media. The statistical agency consults users on a regular basis to find out about the formats of dissemination that they most prefer. Agreements with key users are established for efficient and regular transmission of statistics and data. Technical solutions for access to anonymized data are available. Explicit consideration has been given to trade-offs between accessibility and confidentiality (i.e. level of detail in tables).
17.4 Access to microdata is allowed for research purposes, subject to specific rules and protocols on statistical confidentiality that are posted on the statistical agency’s website The statistical agency controls or monitors the access of researchers to microdata by providing them in a secure environment. The agency needs to have a specific platform for managing release of microdata. Guidelines should be in place on management and use of such data. These may include Data Dissemination and Access Policy, among others

Researchers are regularly consulted about the effectiveness of the microdata access arrangements.

Remote access facilities are available for accessing microdata, with appropriate controls.

65 Requirement Element to be assured Description
17.5 Mechanisms are in place to promote statistical literacy The statistical agencies have a strategy to manage media relationships and maintain regular contact with the media. Communication strategy and Training Plan should address the issue of statistical literacy especially among users of statistics

The statistical agencies arrange regular training and outreach for journalists. The statistical agencies arrange training for students on how to use statistics. The publication of articles on statistical issues, and how statistics should be used properly, is encouraged. 17.6 The statistical agencies have a dedicated focal point that provides support and responds to inquiries from users in a timely manner. There are well-known user support services available to give prompt assistance to users to help them access and interpret the data. The agency needs to establish a customer care center to deal with such inquiries User support services are appropriately staffed to support a wide range of users. 17.7 Users are kept informed about the quality of statistical outputs. Standard quality reports harmonized for the NSOs and as appropriate for the NSS, and tailored for different users’ needs, are defined. Data quality assessment reports need to be made publicly available through website and other media Published statistics are accompanied by standard quality reports, including information on the periodicity of the statistics, data sources, production methods and quality, i.e. about accuracy and reliability, timeliness and punctuality, coherence and comparability, accessibility and clarity. Results from quality assessments or reviews are made public.