Construction and application of Bayesian networks to support decision-making in the water, sanitation and hygiene sector: A case study of SIASAR initiative in Central America
Abstract
The 2030 Agenda includes a dedicated goal on water and sanitation (SDG 6) that sets out to “ensure availability and sustainable management of water and sanitation for all”. SDG 6 expands the MDG focus on drinking water and... [ view full abstract ]
The 2030 Agenda includes a dedicated goal on water and sanitation (SDG 6) that sets out to “ensure availability and sustainable management of water and sanitation for all”. SDG 6 expands the MDG focus on drinking water and sanitation to cover the entire water cycle. A clear lesson from the MDGs is that we cannot manage what we do not measure, and there is little doubt about the role of monitoring and evaluation data in providing the evidence base for decision-making. Against this background, a number of composite indicator frameworks have been implemented to make water and sanitation quality services available and accessible to all, particularly to those most in need. Despite their significance in sector monitoring, indicators are not completely adequate to provide an insight into the complex cause and effect relations within water, sanitation and hygiene (WaSH) issues. The flexibility of Bayesian Networks (Bns) have been exploited to integrate multiple and simultaneous cause-effect or dependence relationships and unravel the linkages between poverty and WaSH services. In consequence, Bns have proved to be effective in project planning and monitoring support. Three major weaknesses however hinder a wider use of this monitoring approach in sector planning: i) an increased data demand, ii) software availability to run the networks, and iii) lack of a systematic methodology to deal with networks generation. In this study, open-databases are exploited and free software “R” is applied. One monitoring initiative is selected as initial case study due to its increasing relevance in Latin America in monitoring the WaSH sector: the Rural Water Supply and Sanitation Information System (SIASAR). On the basis of SIASAR’s conceptual framework, a simple Bn model has been applied to reflect the main issues that determine access to WaSH services. Data from Nicaragua is processed and analysed, since the Government has already carried out and completed a national baseline. The paper discusses about i) the proposed methodology to construct the networks, and ii) the potentiality of BNs in terms of evaluation and planning. It concludes that the proposed methodology represents a contribution to facilitate the use of this tool and that Bns are able to accommodate the complexities of WaSH-related issues. Additionally, they emerge as an effective management tool to support decision-makers in formulating and making informed choices between alternative actions.
Keywords: Bayesian Networks, WaSH, decision-making, planning, SIASAR
Authors
-
Agustí Pérez-Foguet
(Engineering Science and Global Development (EScGD) Research Group, Department of Civil and Environmental Engineering, Civil Engineering School (ETSECCPB), Universitat Politècnica de Catalunya (UPC), Barcelona)
-
David Requejo Castro
(Engineering Science and Global Development (EScGD) Research Group, Department of Civil and Environmental Engineering, Civil Engineering School (ETSECCPB), Universitat Politècnica de Catalunya (UPC), Barcelona)
-
Ricard Giné Garriga
(Engineering Science and Global Development (EScGD) Research Group, Department of Civil and Environmental Engineering, Civil Engineering School (ETSECCPB), Universitat Politècnica de Catalunya (UPC), Barcelona)
-
Gonzalo Martínez Crespo
(Water Global Practice, the World Bank, the World Bank Group, Washington D.C.)
-
Antonio Rodríguez Serrano
(Water Global Practice, the World Bank, the World Bank Group, Washington D.C.)
Topic Area
1a Sustainable development science: fundamental concepts (definitions, fundamental concept
Session
1A-1 » 1a Sustainable development science: fundamental concepts (11:30 - Wednesday, 14th June, SD 703)
Presentation Files
The presenter has not uploaded any presentation files.