Experts discuss key data challenges for hard-to-survey populations and explore solutions in THINK-PA Webinar

The Technical Network on Poverty Analysis (THINK-PA) hosted an engaging webinar titled Beyond the data gap: Unlocking insights to identify and analyse hard-to-survey rural and vulnerable populations. The event brought together leading experts to examine challenges and innovations in data collection for marginalized and hard-to-reach communities.  

The session featured presentations by Brad Edwards (Westat), researchers from the Italian National Institute of Statistics, Valeria de Martino and Nadia Nur, Jeffrey Bloem from International Food Policy Research Institute (IFPRI), and Tim Beatty from the University of California, Davis. Discussions covered difficulties in reaching these populations, recent methodological advancements, and implications for policy.   

Key insights 

During his presentation on “Understanding Hard-to-Survey Populations”, Brad Edwards outlined five dimensions to frame hard-to-survey groups: hard-to-sample, hard-to-identify, hard-to-find or contact, hard-to-persuade, and hard-to-interview. He emphasized that the survey objective significantly influences the level of difficulty and advocated for adaptive approaches such as location-based sampling and community involvement.   

ISTAT researchers presented their work on surveying Roma communities, homeless individuals, and LGBT+ groups in Italy. They highlighted the limitations of traditional sampling methods, such as the "random walk" approach, and stressed the need for mixed-method strategies to improve data accuracy and community trust.   

IFPRI’s Jeffrey Bloem focused on the challenges of collecting data on intermediary actors in agricultural value chains, often overlooked in research. His team introduced a network-driven sampling approach that tracks transaction linkages in real time, revealing inefficiencies and opportunities for policy intervention in Uganda and Bangladesh. 

Tim Beatty closed the session with an analysis of farmworker invisibility in labour data systems, particularly in the United States of America. He highlighted the economic and health risks faced by these workers, particularly regarding heat exposure and workplace injuries. His research team is leveraging machine learning models and mobile location data to fill these gaps and inform policy discussions on labour protections.   

Looking ahead   

The webinar underscored the importance of methodological innovation in surveying hard-to-reach populations. Key takeaways included: 

• Hybrid sampling techniques, such as network-driven and dual-frame sampling, offer new ways to capture hidden groups. 
• Machine learning and big data can enhance existing datasets and reduce reliance on costly small-scale surveys. 
• Strong engagement with local communities remains essential to improving study design, participation and data quality. 

As data collection methodologies continue to evolve, these insights will play a crucial role in shaping future research and policy efforts to ensure inclusive agrifood systems and effective poverty analysis.