BACK TO SCHEDULE REGULAR

Session 7

Building Quality Through Standards

3 June 2026
14:15 – 15:45
ŠIBENIK III

Presentation title
Scientific Foundations for Processing and Using Data in Official Statistics
European statistics must provide an accurate and reliable representation of reality.

Read more Read less

This is emphasized in the regulation (EC) No 223/2009 of the European Parliament and of the Council. Scientific methods are considered essential to ensure that statistics “measure as faithfully, accurately and consistently as possible the reality they are designed to represent.” Traditionally, the design of statistical surveys and the compilation of statistics based on appropriately collected data have relied on well-established statistical science.

However, when data originally collected for non-statistical purposes is repurposed for official statistics, it is less clear which approaches and strategies can be regarded as scientifically sound.

In 2025, the Development Unit at Statistics Sweden’s Department of Data Management initiated a project to define criteria ensuring that the collection, processing, and use of data for official statistics adhere to scientific principles and meet key standards. The underlying premise is that a scientific approach and methodological transparency are critical for building trust in official statistics.

The project resulted in a set of key characteristics describing data-related processes that follow scientific principles. Each characteristic is explained in detail to raise awareness and provide a foundation for identifying and managing uncertainty — whether stemming from data properties or from the methods used in processing and applying data in statistical production.

Main author / Presenter
Natalie Jansson, Kristina Strandberg
Statistics Sweden

Read more Read less

Kristina Strandberg holds a Master of Science in Statistics from the University of New Mexico, earned in 2003. After completing her degree, she pursued competitive cross-country skiing for several years, winning titles such as the U.S. National Championships. Returning to her native Sweden, Kristina joined Statistics Sweden (SCB), where she started her tenure a methodologist focusing on the Consumer Price Index and foreign trade. Her interest in leadership and strategic development later led her to management roles. Today, Kristina is Head of Section for the methodologists within the Development Unit of the Department of Data Management. Her team is responsible for ensuring that methodological aspects are integrated into the management and development of data collection and data processing, safeguarding quality and scientific rigor in official statistics.

CO-AUTHOR:

Presentation title
"Innovation in Practice: Adapting the UN National Quality Assurance Framework for NSIs’ Implementation"
International quality frameworks provide essential guidance for quality management at NSIs, yet their operationalisation requires substantial adaptation respecting local contexts while maintaining alignment with underlying quality principles.

Read more Read less

This paper examines the experience of the National Statistics Office of Mongolia in implementing the UN National Quality Assurance Framework (NQAF) Self-Assessment Checklist to assess the quality of its statistical processes, revealing fundamental tensions between international standards and local operational realities.

The implementation revealed five critical challenges: ambiguity in assessment objectives, with the UN checklist attempting to serve both system-wide and process-specific diagnostic purposes; redundancies and excessive generality in the 356 "elements to be assured"; impractical organization of data collection when a central team attempts to conduct an organizational assessment across the board; insufficient sensitivity of the three-point compliance scale; and technology infrastructure limitations.

To address these challenges, the NSO of Mongolia introduced significant methodological innovations. It developed process-specific assessment instruments tailored to six distinct process types, ensuring the administration of relevant questions for each type while eliminating irrelevant elements. The assessment was conducted at the element level rather than at the requirements level, providing the diagnostic precision necessary for improvement planning. A five-point scoring scale replaced the original three-point scale, resulting in meaningful differentiation of quality levels and identification of priorities for improvement. A decentralised approach was adopted during implementation, with process owners conducting initial assessments subject to management review, balancing detailed operational knowledge with objectivity concerns. An open-source software tool replaced Excel, enabling real-time validation, centralised data management, and longitudinal tracking capabilities.

A cyclical quality assessment strategy was designed to embed quality management as an organisational practice rather than a one-off exercise. The strategy foresees conducting a self-assessment of all statistical processes every three years, establishing benchmarks for measuring progress over time. Three priority processes are selected for annual in-depth reviews based on assessment scores of the processes and the strategic importance and visibility of their outputs. These in-depth assessments engage external experts to provide independent perspectives, introduce international good practices, and mitigate self-assessment subjectivity. Each in-depth review results in detailed improvement plans, specifying objectives, milestones, timelines, and concrete actions addressing identified weaknesses. This cyclical approach balances comprehensive coverage against resource constraints.

The methodology innovations documented here provide a replicable model for statistical agencies worldwide seeking to strengthen quality assurance operations while managing resource constraints, transforming international frameworks from theoretical constructs into practical management tools that genuinely support quality improvement

Main author / Presenter
Pietro Gennari
International Statistical Institute

Read more Read less

Pietro Gennari is currently Professor at the Master in Statistical Methods and Applications, Sapienza University of Rome, Editor-in-Chief of the Statistical Journal of the IAOS and Senior Consultant in Official Statistics and SDG monitoring. With over 35 years of professional experience in the main areas of Official Statistics, managing large statistical programmes at national, regional and international levels, he has acquired an in-depth knowledge of the methods and data sources used in a variety of statistical domains and extensive experience in providing support to strengthen the statistical capacity of developing countries. From December 2008 to July 2023, he worked as Chief Statistician of FAO. During 2014-2017, he also served as Chair of the Chief Statisticians of the UN system. Before joining FAO, he was the Director of the Statistics Division at the UN Regional Commission for Asia-Pacific, and he served for 15 years at ISTAT with progressively higher managerial responsibilities

CO-AUTHOR:

Presentation title
Standardization of Data Processing to Improve Quality in Official Statistics
Official Statistics in Germany is undergoing a comprehensive modernization that is redefining the processes of the data processing and particularly plausibility checking as a central component of data processing.

Read more Read less

The need for this modernization arises from a long-standing trade-off between increasingly complex data sources, high requirements for timeliness and the need to ensure reliable and high-quality results. Existing, partly historically evolved and heterogeneous plausibility checking procedures are reaching conceptual and organizational limits. This applies to both the diversity of individual workflows and subject-specific solutions within the statistical system and the technological limitations of existing systems.

Overall, official statistics pursue the goal of developing data processing as a future-proof, integrated and largely automated component of the whole statistical production process. The modernization addresses subject-matter, organizational and structural challenges alike in order to enable a long-term transition toward more efficient, consistent and high-quality data production. The modernization of data processing and plausibility checking processes therefore relies on consistent standardization and harmonization of workflows, involved roles and subject-matter requirements. The aim is to create integrated, transparent and efficiently controllable to-be processes that can be applied across all statistics and that enable a high degree of comparability. The basis lies in a comprehensive standard process model within GMAS Phase 5, the so-called data processing map and the corresponding BPMN model, which are based on international standards such as GSDEM, GSBPM and GSIM. These models enable describing the functional core components of plausibility checking processes - error detection, error handling, error correction and quality assurance - in a modular, transparent and reusable manner. The standard processes of the data processing are systematically represented in the data processing map through a combination of hierarchical process structures (sub-processes, activities, sub-activities) and supporting process functionalities, detailed textual process descriptions, workflow diagrams and definition tables.

An organizational enhancement of the data processing map and the corresponding BPMN model ensures consistent and standardized implementation of the statistical subject-matter to-be processes and their integration into future target architectures, such as the Statistical Data Ware House. The modernization of plausibility checking processes thus represents a key building block for aligning statistical production in the long term toward greater efficiency, consistency and robust quality.

This paper introduces the German data processing map and provides a glimpse of a corresponding BPMN model as tools for standardization and modernization of the data processing phase of a statistical production process for a sustainable data quality improvement.

Main author / Presenter
Arijana Amina Ramic
Federal Statistical Office of Germany

Read more Read less

Arijana Amina Ramic studied mathematics and information technologies and has been working in official statistics for a total of 20 years. She has been working for the past 10 years at the Federal Statistical Office of Germany, where she is mainly responsible for the improvement of the quality of the statistical products through the improvement and modernization of the statistical processes and tools. Thus, she is jointly responsible for the standardization, automation and optimization of various subprocesses of the whole statistical production process as well as of the tools used based on the new conceptual ideas and results of digital assessments in decentralized statistics. She is included through different aspects and projects in the development of the Statistical Data Ware House as well as in the development of the standardized statistical production process within the German Statistical System.

Presentation title
Linking the UN National Quality Assurance Framework and the Generic Statistical Business Process Model
In 2019, the UN Expert Group (EG) on National Quality Assurance Frameworks (NQAF) issued the “United Nations National Quality Assurance Frameworks Manual for Official Statistics” (https://unstats.un.org/UNSDWebsite/data-quality/user-manual), which includes the generic UN NQAF.

Read more Read less

Afterwards, the EG delivered several tools to support and facilitate the implementation of a NQAF, such as the Self-assessment Checklist, the Roadmap for implementation. , and, more recently, the Maturity Model on Quality Culture in Official Statistics and the Module for Quality Assurance when using Administrative and Other Data Sources to produce Official Statistics. Such tools proved to be very effective.

Since the beginning of 2025, the EG has been working on linking the UNECE Generic Statistical Business Process Model (GSBPM) with the UN NQAF. The GSBPM is often referred to as a powerful supporting tool for achieving quality improvements, but clear guidelines on how to proceed are still not available.

In order to carry on this activity, a subgroup of the EG was set up, co-chaired by Istat, Italy and DANE, Colombia, with a twofold objective: first, to develop the GSBPM overarching process of quality management and, second, to develop guidelines for the use of GSBPM to support the implementation of the NQAF.

The first output will be a set of good practices and quality indicators for the GSBPM sub-processes. The work started with a revision of the 2017 UNECE document on Quality indicators for GSBPM (https://statswiki.unece.org/download/attachments/185794796/Quality%20Indicators%20for%20the%20GSBPM%20-%20For%20Statistics%20derived%20from%20Surveys%20and%20Administrative%20Data%20Sources_Final.pdf?api=v2). The quality indicators have been classified into qualitative and quantitative ones and assigned different priorities. The qualitative ones have been transformed into good practices. In contrast, the quantitative high-priority ones have been equipped with formulas and further remarks, e.g., if they refer to input, process, or output quality. Both the good practices and the quality indicators have been mapped with the Principles and Requirements of the UN NQAF version 2019 and with the GSBPM sub-processes version 5.2. This deliverable should provide concrete guidelines for the implementation of quality assurance at the level of the statistical process.

The second output will be guidelines and, if possible, a roadmap on how to use the GSBPM to support NQAF implementation at a more cross-cutting level. A workshop and a survey on this topic were conducted to collect information on existing practices in different countries. Those will be the main inputs for the guidelines. The work of the subgroup will be presented with the aim of collecting further feedback and suggestions for improvement.

Main author / Presenter
Giorgia Simeoni
Istat

Read more Read less

Giorgia Simeoni works for Istat, Italian National Statistical Institute, since 2001, and has been appointed as Istat Quality manager in September 2020. She has always worked in the field of quality and reference metadata, has been responsible for implementing quality reporting at Istat. She is involved in European and international quality and standards related working groups, like the UN Expert Group on National Quality Assurance Framework and the UNECE Supporting Standards Group.In recent years has worked on the development and implementation of quality frameworks for multisource statistics and for statistics based on new data sources. She has also a relevant experience in cooperation projects related to quality topics and is also a trainer in ESTP courses. She is graduated in statistics, demographic and social sciences and holds a master degree in sources, tools and methods for social research.

CO-AUTHOR:

Presentation title
Incorporating UNECE/ESS standards and control their application using ISO-9001.
Over the past two years, with the arrival of the Data Office guys, the Basque Statistical Institute (EUSTAT) have shared ideas and worked together to explore how we could access to government’s data without causing “collisions” in the Informational layer of the Basque Country’s administration.

Read more Read less

The tools and paradigms currently used for data governance have made us realize two things:

• Part of the work we do today will no longer be part of our responsibilities, so we will need to specialize in the purely statistical domain and give space to the new data owners, custodians, analysts, etc.

• Full convergence with the Data Office’s information systems is not completely possible, and that we are moving toward a dual structure: one government-level information system and one statistical system. This means that automated data exchange will become mandatory if we want to remain performant (exchange not only with external entities outside the Basque Government, but in a near future also internally within government), along with ensuring its confidentiality.

Given that, it was clear that we needed to adapt both our working methods and our Information systems. In this paper, we will focus on the first one, although we expect that doing this the right way will also facilitate changes to our IT systems.

EUSTAT had been operating with the same process map for more than 15 years, originally implemented under the EFQM model, with only minor adjustments over time. This process map, and the tools derived from it (overall the statistical program’s “technical project”) is under a completely redesign to align with the standard GSBPM + GAMSO model.

In this way, we aim to use a common vocabulary shared with other statistical institutes, enabling us to store information in a standardized manner (through the use of GSIM + COOS standards) and leverage currently available Open-Source tools.

We all know that implementing effective process-based management in an organization is a comprehensive change project that impacts organizational culture (particularly interpersonal relationships), resources (especially technological tools), and systems (notably management procedures). Therefore, we started from a desired “Vision for Change”: adapting to the consequences of the Data Office’s arrival, and we are attempting to implement this process-based approach gradually, while avoiding statistical program’s “stovepipe organization”.

To ensure that everything is being implemented as designed, we plan to monitor it by incorporating statistical programs into the ISO-9001 audit process and expanding its scope.

Main author / Presenter
Jesus Nieto Gonzalez
EUSTAT - Basque Statistics Institute

Read more Read less

Computer engineer by training and data manager by passion, with extensive experience both in the business sector and in the public administration, where he has been working for more than 15 years, the most recent of them at EUSTAT. In addition to holding various positions, all related to the development of IT projects from different perspectives, he has served as a key reference for personal data protection for the Data Protection Officer of the Basque Government and as EUSTAT’s security officer. Currently, he is responsible for EUSTAT’s technological-digital transformation and has led several projects related to data governance developed within EUSTAT.