Revealing the Relational Mechanisms of Research for Development Through Social Network Analysis.

Achieving impact through research for development programmes (R4D) requires engagement with diverse stakeholders across the research, development and policy divides. Understanding how such programmes support the emergence of outcomes, therefore, requires a focus on the relational aspects of engagement and collaboration. Increasingly, evaluation of large research collaborations is employing social network analysis (SNA), making use of its relational view of causation. In this paper, we use three applications of SNA within similar large R4D programmes, through our work within evaluation of three Interidsiplinary Hubs of the Global Challenges Research Fund, to explore its potential as an evaluation method. Our comparative analysis shows that SNA can uncover the structural dimensions of interactions within R4D programmes and enable learning about how networks evolve through time. We reflect on common challenges across the cases including navigating different forms of bias that result from incomplete network data, multiple interpretations across scales, and the challenges of making causal inference and related ethical dilemmas. We conclude with lessons on the methodological and operational dimensions of using SNA within monitoring, evaluation and learning (MEL) systems that aim to support both learning and accountability. Supplementary Information: The online version contains supplementary material available at 10.1057/s41287-023-00576-y.

Pour que les programmes de recherche pour le développement (R4D ou Research for Developmement en anglais) aient un impact, il faut un engagement entre diverses parties prenantes dans les domaines de la recherche, du développement et des politiques. Il est nécessaire de se concentrer sur les aspects relationnels de l'engagement et de la collaboration si l'on souhaite comprendre la façon dont ce type de programme permet l'émergence de résultats. L'évaluation des grands consortia de recherche utilise de plus en plus fréquemment l'analyse des réseaux sociaux (SNA ou social network analysis en anglais) en appliquant sa vision relationnelle de la causalité. Dans cet article, en vue d'explorer son potentiel en tant que méthode d'évaluation, nous utilisons trois applications d'analyse des réseaux sociaux au sein de grands programmes R4D similaires dans le cadre de notre travail d'évaluation de trois pôles interdisciplinaires du Fonds de recherche sur les défis mondiaux. Notre analyse comparative montre que l'analyse des réseaux sociaux peut révéler les dimensions structurelles des interactions au sein de ces programmes et permettre d'apprendre comment les réseaux évoluent dans le temps. Nous menons une réflexion quant aux défis communs qui émanent de ces cas, y compris la gestion de différentes formes de biais qui résultent de données de réseau incomplètes, de multiples interprétations sur des échelles différentes et les défis liés au fait d'établir une inférence causale et les dilemmes éthiques connexes. Nous concluons par des leçons sur les dimensions méthodologiques et opérationnelles de l'utilisation de l'analyse des réseaux sociaux dans les systèmes de suivi, d'évaluation et d'apprentissage (SEA) qui visent à soutenir à la fois l'apprentissage et la redevabilité.

Towards the Three Dimensions of Sustainability for International Research Team Collaboration: Learnings from the Sustainable and Healthy Food Systems Research Programme.

This paper highlights the potential for learning and virtual collaboration in international research teams to contribute towards sustainability goals. Previous research confirmed the environmental benefits of carbon savings from international virtual conferences. This paper adds the social and economic dimensions by using a combination of qualitative and quantitative methods to measure the constraints and benefits for personal development, economic costs, efficiency and team learning of holding international virtual conferences (VCs). Using the Sustainable and Healthy Food Systems (SHEFS) research programme as a case study, we analysed VC participant survey data to identify strengths, weaknesses, opportunities, and threats of VCs. We estimated 'saved' GHG emissions, costs, and time, of using VCs as an alternative for a planned in-person meeting in Chennai, India. Hosting VCs reduced North-South, gender, and researcher inclusivity concerns, financial and travelling time costs, and substantially reduced emissions. For one international meeting with 107 participants, changing to a virtual format reduced the per capita GHG emissions to half the annual global average, and avoided 60% of travel costs. The benefits of VCs outweighed weaknesses. The main strengths were inclusivity and access, with 20% more early/mid-career researchers attending. This study identified opportunities for international research partnerships to mitigate their carbon footprint (environmental benefit) and enhance inclusivity of early/mid-career, women and Global South participants (social benefit), whilst continuing to deliver effective collaborative research meetings (economic benefit). In doing so, we present a holistic view of sustainability opportunities for virtual collaboration.

Comparison of Alternative Meat Inspection Regimes for Pigs From Non-Controlled Housing - Considering the Cost of Error.

Denmark has not had cases of bovine tuberculosis (bovTB) for more than 30 years but is obliged by trade agreements to undertake traditional meat inspection (TMI) of finisher pigs from non-controlled housing to detect bovTB. TMI is associated with higher probability of detecting bovTB but is also more costly than visual-only inspection (VOI). To identify whether VOI should replace TMI of finisher pigs from non-controlled housing, the cost of error - defined here as probability of overlooking infection and associated economic costs - should be assessed and compared with surveillance costs. First, a scenario tree model was set up to assess the ability of detecting bovTB in an infected herd (HSe) calculated for three within-herd prevalences, WHP (1, 5 and 10%), for four different surveillance scenarios (TMI and VOI with or without serological test, respectively). HSe was calculated for six consecutive 4-week surveillance periods until predicted bovTB detection (considered high-risk periods HRP). 1-HSe was probability of missing all positives by each HRP. Next, probability of spread of infection, Pspread, and number of infected animals moved were calculated for each HRP. Costs caused by overlooking bovTB were calculated taking into account Pspread , 1-HSe, eradication costs, and trade impact. Finally, the average annual costs were calculated by adding surveillance costs and assuming one incursion of bovTB in either 1, 10 or 30 years. Input parameters were based on slaughterhouse statistics, literature and expert opinion. Herd sensitivity increased by high-risk period and within-herd prevalence. Assuming WHP=5%, HSe reached median 90% by 2nd HRP for TMI, whereas for VOI this would happen after 6th HRP. Serology had limited impact on HSe. The higher the probability of infection, the higher the probability of detection and spread. TMI resulted in lowest average annual costs, if one incursion of bovTB was expected every year. However, when assuming one introduction in 10 or 30 years, VOI resulted in lowest average costs. It may be more cost-effective to focus on imported high-risk animals coming into contact with Danish livestock, instead of using TMI as surveillance on all pigs from non-controlled housing.