UNECE Modernisation of Official Statistics

➡️ New Paper: Strengthening Employer Branding in NSOs In today’s competitive labor market, National Statistical Offices (NSOs) face growing challenges in attracting and retaining top talent, especially in IT and data science. A strong employer brand is essential. Led by Renata Nowicka of Statistics Poland, the new paper “Employer Branding in NSOs” from the HLG-MOS Capabilities and Communication Group explore the key elements that influence the brand management and assess current employer branding practices and identify areas for growth through analysis of survey result. 👉 Key strategies for NSOs to strength employer branding include: - Develop a strategic employer branding plan with clear goals - Establish a compelling employee value proposition (EVP) - Invest in digital marketing and modern communication channels with regularly assessing employer branding initiatives to ensure continuous improvement - Use the generic growth model (GGM) (https://lnkd.in/eXD7x2w3) for self-assessment and improvement Employer branding is more than just an HR initiative - it’s a strategic necessity that requires collaboration between HR, communication, and leadership. ✨ Let’s share best practices and build stronger, more attractive NSOs! Read the full paper here: https://lnkd.in/efJADd5H #EmployerBranding #HR #TalentAttraction #CapComm #ModernStats #HLGMOS

📢 New Release: Chapter on Using, Implementing, and Developing Generative AI in Official Statistics A new chapter from the forthcoming report, Generative AI in Official Statistics, is now available. This is the second chapter in the report, set to be fully published in mid-2025, focusing on practical applications, implementation strategies, and technical considerations for integrating generative AI into statistical organizations. 🌟 Key highlights from the chapter: - Identifying application areas - How generative AI supports the statistical production process, from data collection to dissemination. - Effective prompt engineering - Best practices to optimize AI-generated outputs. - Building generative AI solutions - Key technical and infrastructure considerations for secure and efficient AI implementation. - Boosting AI performance - Techniques such as retrieval-augmented generation (RAG), fine-tuning, and agent-based frameworks to improve accuracy and reliability. 🔑 This chapter provides real-world examples and recommendations to guide statistical organizations in effectively leveraging generative AI while maintaining high standards of accuracy, transparency, and security. Read the whole chapter here: https://lnkd.in/e9r_uhWF 🤲 We value your insights! Share your thoughts and feedback on this chapter in the comments below or via https://lnkd.in/ewX_zKQh 🔜 Stay tuned for more chapters in the coming months as we continue to explore the evolving role of generative AI in official statistics. #GenerativeAI #OfficialStatistics #ModenStats #HLGMOS

💡 HLG-MOS Annual Meeting: Key Takeaways Each year, HLG-MOS holds a meeting at the margins of the UN Statistical Commission to review progress and set priorities for its future work programmes. Chaired by Angelique Berg (Centraal Bureau voor de Statistiek, HLG-MOS Chair) with Phillip Gould (Australian Bureau of Statistics, Executive Board Co-chair) presenting on results from 2024 and plans for this year, the meeting at the 56th Statistical Commission discussed challenges and strategic direction for 2025 and beyond. 🤔 What are the key points raised at the meeting? - Transition to multi-source while improving primary collection: A few years ago, focus was on gaining access to privately held data. The challenge is increasingly how to effectively leverage multi-source statistics while maintaining the value of primary data collection. The two new HLG-MOS projects on multi-source and survey non-response are very timely and complement each other well. - Role of NSOs in a broad government data system: Other government departments are already linking their data for operational purposes which then can lead to data integration for analytical purposes as well. How NSOs should position themselves and promote the efficient public sector-wide integrated data system? - Generative AI and pathways for NSOs: We should work on two key approaches - ensuring our data is findable and accurately reflected in widely used foundational AI models, while also providing reliable tools that generate responses based on curated, high-quality data. - Going slow to go faster: The future of NSOs is about becoming faster and more responsive, but building lasting relationships takes time. We must be prepared for long-term investment for the engagement with public and communities, approaching it on their terms, not the government's. - Partnership with stakeholders: beyond legal agreements, building strong relationship with data providers is important factor to ensure data quality. Data protection is one area of partnership where engagement with national privacy authorities play a crucial role. Additionally, partnering with private companies can drive the development of AI and privacy-enhancing solutions. Read more about HLG-MOS work of 2024 and plan for 2025 here: https://lnkd.in/eAeW49su #HLGMOS #ModernStats

🚀 New HLG-MOS Release: Statistical Open Source Software We are excited to announce the release of the Statistical Open Source Software - Charter and Report! This report aims to help statistical organizations looking to navigate the opportunities and challenges of open source software (OSS) adoption. 🔍 What’s Inside? - OSS Principles - Fostering a common understanding of the expected behaviour and practices when using OSS which ultimately can help the statistical community to collaborate more effectively. - Strengths & Weakness of Open Source - Understanding the opportunities and challenges for statistical organizations when adopting OSS - Real-World Case Studies - How national and international statistical organizations are implementing OSS. - Practical Recommendations - Key considerations arising when adopting OSS from licensing, knowledge building and security. 💡 Why This Matters Open source is more than just making source code freely available - it’s a collaborative shift that requires cultural change, strategic investment, and governance frameworks. As statistical organizations modernize their production, OSS is playing a critical role in driving digital transformation and fostering collaboration across the global statistical community. We appreciate Centraal Bureau voor de Statistiek (Statistics Netherland), Istat, OECD - OCDE and Statistical Office of the Republic of Serbia for case studies, and Carlo Vaccari for leading the project!  📖 Read the full report here: https://lnkd.in/eiH9R7Au 🌍 Explore more OSS initiatives: https://lnkd.in/eFHV9dqM

✨ Generative AI Workshop - We Need Your Insights on Policy, Partnerships & Industry Collaboration! The Generative AI and Official Statistics Workshop (12-14 May 2025, Geneva) is shaping up to be an exciting platform for discussion and exchange on various application of generative AI in context of official statistics. To ensure a well-rounded conversation, we are particularly looking for contributions in following areas: 🔹 Governance, risk management & strategies - How can we ensure responsible adoption of generative AI in official statistics? What frameworks and strategies are needed? How can organizations prepare for, regulate, and strategically integrate Generative AI? 🔹 Building Partnerships - Are you collaborating with private companies, academia, or other agencies on generative AI initiatives? Share your experiences in forging partnerships and navigating cross-sector collaboration. 🔹 Private Sector Perspectives - Are you a private company working with official statistics? We invite you to share your insights, challenges, and solutions in applying AI-driven technologies to statistical processes. 💡 We don’t only seek stories of success! Whether your experience was a breakthrough, a challenge, or a valuable lesson learned - we want to hear from you. Your insights can help shape the conversation on how Generative AI can be effectively used in official statistics. 🔗 Submit your abstract by 5 March 2025: https://lnkd.in/eP8v3Bdg #GenerativeAI #ModernStats

💡Data from national statistics offices (NSOs) provide the foundation for evidence-based decision and policy-making for our society. Similarly, HR data analytics can offer valuable, data-driven insights to support strategic planning and resource allocation within NSOs and help them optimize their workforce and operations. ✅Led by Sarah Johnston-Way of Statistics Canada | Statistique Canada the paper “Enhancing NSOs through HR Analytics” presents the value proposition of HR data analytics, concrete examples of how it is implemented in various organizations, and some key considerations for adopting HR analytics in statistical organizations. ⚖️ As NSOs face increasing budget pressures, balancing between priorities, and adopting rapidly evolving technologies, this paper will provide a valuable resource for those seeking to adopt evidence-based approaches to human resource management. ➡️ The full report is available here: https://lnkd.in/erbkhXDm #ModernStats #HLGMOS #CapabilityCommunicationGroup

🚀 Launch of a new HLG-MOS Project - A Roadmap to Multi-Source Statistics (RAMSES)! 💡 Why this project? Data integration and linkage have become essential for statistical organizations, enabling them to provide richer insights and timely information to address complex challenges in an ever-evolving world. Over the past decade, the topic has evolved. Organizations are now striving to move beyond one-off data integration solutions toward a more systematic approach. 💡 What is the objectives of the project? The project will focus on, rather than how to integrate data, but how to build a strong foundation to make data integration more effective and efficient, thus allowing the ultimate end result - the production of multi-source statistics - a standard mode of statistical production. The project aims to develop a Handbook on Multi-Source Statistics, addressing key areas such as: - Holistic transition and rethinking of processes design - Interoperability and standardization in a practical sense; - Ethics and social acceptability in context of multi-source statistics; - A framework to understand various quality aspects and communication of the quality with users. Additionally, we plan to test the handbook through concreate use cases. 👉 You can find more information on the project proposal here: https://lnkd.in/eiNckpyX 🤲 Want to join us on this exciting journey? Fill out the form and we will get in touch: https://lnkd.in/eqfE2dSt #ModernStats #HLGMOS #MultiSourceStatistics #DataIntegration #RAMSES

📢 Announcing the Launch of the ASCENT Project We are pleased to introduce a new HLG-MOS project of 2025 - Advanced Survey Cost-Effectiveness with Nonresponse Treatment (ASCENT) Project, an initiative focused on addressing nonresponse bias and enhancing survey processes through practical, implementation-driven methodologies. Declining response rates and increasing budgetary pressures underscore the urgent need for sustainable and efficient survey methodologies to safeguard the integrity of official statistics. The project aims to deliver guidelines to mitigate nonresponse bias and optimize survey processes by leveraging advanced sampling and collection methods. The project is not just about theory - it is about turning the research and innovations NSOs are already testing into a shared, actionable framework. We will focus on: 🔹 Responsive Design and Multi-Mode Approaches – Adjusting survey processes dynamically to improve efficiency and quality. 🔹 Subsampling of Nonrespondents – Implementing targeted follow-ups to enhance data accuracy while managing costs. 🔹 Advanced Post-Collection Weighting – Applying refined techniques to correct for nonresponse bias and support high-quality statistical outputs. The project will culminate in a standardized report, offering practical tools and methodologies designed primarily for and the academic community. 👉 You can find more information on the project proposal here: https://lnkd.in/eHuK6mXf 💡 Interested in joining this initiative? If you’d like to contribute or learn more, we invite you to fill out the form here: https://lnkd.in/e4R-myrk Your expertise and participation can help us shape a more sustainable and effective future for data collection. #HLGMOS #ModenStats #Ascent

🚀 New Release: Use of Generative AI for Communication in Statistical Organizations. The use of AI in statistical organizations is on the rise, and generative AI, with its ability to produce text, images, and videos from simple cues offers a unique opportunity to revolutionize communication. ✨ As communication becomes increasingly critical for producers of official statistics, generative AI can: - Streamline routine tasks - Create highly tailored materials for specific demographic groups - Strengthen connections with target audiences - Boost public engagement 📱 As seen in the case of social media, technological advancements can reshape how statistical organizations communicate with the public. However, given that communication is at the forefront with users and the public interacting directly, official statistics producers must remain vigilant about emerging risks and challenges when adopting new technology. 🎯 Led by Janice Keenan (co-chair, @Statistics Canada) and Bilyana Bogdanova (co-chair, Bank for International Settlements – BIS), this paper explores: ✅ Opportunities offered by generative AI through real use cases ✅ Risks are more pronounced when used for communication ✅ Challenges for statistical organizations adopting new technologies Read more here: https://lnkd.in/eWxhSRsr #ModernStats #HLG2024 #HLGMOS #AIforCommunication

📢 Release of a Chapter from the Report on Generative AI in Official Statistics A chapter from the forthcoming report, Generative Artificial Intelligence in Official Statistics, set to be fully published in mid-2025, is now available. This is the fourth chapter in the report’s sequence but the first to be released, with additional chapters to follow in the coming months. Check out the chapter here: https://lnkd.in/eTHyGmma 🌟 This chapter explores the risks and mitigation strategies for integrating Generative AI into institutions responsible for official statistics. Key topics include: ▫️Risks to transparency, traceability, and accuracy due to the stochastic and opaque nature of Generative AI. ▫️Data privacy, legal, and ethical challenges, including potential biases and compliance concerns. ▫️Operational and security risks, such as cybersecurity threats and over-reliance on external resources. ▫️Environmental sustainability considerations amidst rapidly evolving regulatory landscapes. ▫️The hidden complexities of Generative AI and their implications for official statistics. 🔑 The chapter outlines technical, policy, and governance approaches, as potential mitigation strategies emphasizing: ▫️Secure development and use of generative AI. ▫️Open-source AI to enhance reproducibility and transparency. ▫️Tailored evaluation and auditing frameworks for safety, legal, and ethical compliance. 📣 Feedback Welcome! Your perspective is important. Share your thoughts on this chapter via this form: https://lnkd.in/ewX_zKQh 🔜 Additional chapters will be released in the coming months. Stay informed to explore further insights on generative AI and its role in official statistics. #HLGMOS #ModernStats #GenerativeAI