ABSTRACT Issue Addressed Social network analysis (SNA) is a novel approach to investigating food security systems change in regional and remote communities. With short‐term, siloed food security solutions often proving ineffective, this study used SNA to visualise how organisations across a vast area in Western Australia (WA) were working in partnership to support systems change and identify where these partnerships could be strengthened. Methods One hundred and one food security initiative leaders, representing 148 initiatives, participated in semi‐structured interviews. The number and type of regional partnerships were measured, and region‐based partnership network typologies determined. Results Most regions’ partnership networks were ‘Scattered fragments’ models, indicating immature and disconnected partnership networks. The Goldfields region represented a ‘Hub‐and‐spoke’ network, while the Kimberley partnership network was transitioning from ‘Hub and Spoke’ to a more robust ‘Multi‐hub’ model. The application of a theoretical framework based on SNA yielded practical insights into the functional nature of these remote networks and increased policymakers', practitioners' and researchers’ understanding of how they could be leveraged to improve food security, with practical implications for rural public health outcomes. Conclusion Recommendations to strengthen partnership networks included supporting unconnected organisations in each region to connect into their partnership network (if desired). Undertaking targeted partner ‘knitting’ by identifying potential network weavers and connecting partners is recommended. Supporting each region's network to transition across network stages is critical to ensure robust, sustained collaboration, and optimise health outcomes for all community members. So What? This study uncovered which organisations were most influential in information dissemination in rural, regional, and remote Australia and has identified less mature regions which require more equitable investment and resource allocation.
Godrich et al. (Thu,) studied this question.