Data-powered positive deviance: Combining traditional and non-traditional data to identify and characterise development-related outperformers. (2022)

• Record Type: Journal Article
• Authors: Albanna, Basma; Heeks, Richard; Pawelke, Andreas; Boy, Jeremy; Handl, Julia; Gluecker, Andreas
• Is Part Of: Development engineering. Volume 7(2022)
• Publication Date: 2022
• Subjects: International development -- Positive deviance -- Non-traditional data -- Mixed methods -- Data science -- High performers -- Global south; Technical assistance -- Periodicals; Economic development -- Periodicals; Engineering -- Social aspects -- Periodicals; 338.9105
• DOI: 10.1016/j.deveng.2021.100090
• Languages: English
• ISSNs: 2352-7285

Abstract

Abstract: The positive deviance approach in international development scales practices and strategies of positively-deviant individuals and groups: those who are able to achieve significantly better development outcomes than their peers despite having similar resources and challenges. This approach relies mainly on traditional data sources (e.g. surveys and interviews) for identifying those positive deviants and for discovering their successful solutions. The growing availability of non-traditional digital data (e.g. from remote sensing and mobile phones) relating to individuals, communities and spaces enables data innovation opportunities for positive deviance. Such datasets can identify deviance at geographic and temporal scales that were not possible before. But guidance is needed on how this new data can be employed in the positive deviance approach, and how it can be combined with more traditional data to gain deeper, more meaningful, and context-aware insights. This paper presents such guidance through a data-powered method that combines both traditional and non-traditional data to identify and understand positive deviance in new ways and domains. This method has been developed iteratively through six development projects covering five different domains – sustainable cattle ranching, agricultural productivity, rangeland management, research performance, crime control – with global and local development partners in six countries. The projects combine different types of non-traditional data with official statistics, administrative data and interviews. Here, we describe a structured method for data-powered positive deviance developed from the experience of these projects, and we reflect on lessons learned. We hope to encourage and guide greater use of this new method; enabling development practitioners to make more effective use of the non-traditional digital datasets that are increasingly available.