Leveraging the Positive Deviance Approach using Big Data

  • Basma Mahmoud

Student thesis: Phd


This thesis presents a method that combines non-traditional digital data, namely big data, and traditional data to identify and characterise outperformers in development-related challenges. It builds on the “Positive Deviance” (PD) approach for development, which is based on the observation that in every population there are individuals or communities who, despite facing similar challenges and limitations, achieve better results than their peers. This approach focuses on these outliers (or positive deviants) in order to discover unusual practices and strategies that successfully solve complex problems ¬– particularly where conventional solutions failed. In order to build this method, I first conducted a systematic literature review to identify the opportunities and challenges of using big data in the positive deviance approach and outlined a preliminary analytical framework that could guide the use of such data in the PD approach. Following that, I tested and validated this framework in multiple case studies which supported the iterative development and refinement of a method I refer to as “Data Powered Positive Deviance” (DPPD). In the first case study, I used online data along with traditional data sources, such as surveys and interviews, to identify and characterise Egyptian information system researchers who were able to outperform their peers in publication outcomes. I applied the same framework during a six-month fellowship at the United Nations Pulse Lab Jakarta, to identify and validate positively deviant rice-growing villages in Indonesia using official statistics and administrative data along with earth observation big data. This fellowship was part of a global initiative collaboratively created by the GIZ data lab, UNDP accelerator labs, Pulse Lab Jakarta and the University of Manchester Centre for digital development. It builds and scales the DPPD method developed in this study, to see if and how we might use big data-based positive deviance to tackle development challenges. In addition to the Indonesia project, four action research projects were implemented as part of this initiative to identify and understand: farmers achieving higher than usual cereal crop productivity in Niger; cattle farmers in Ecuador who are deforesting below average rates; public spaces in Mexico City where women are safest; and communities in Somalia which are able to preserve their rangelands despite the frequent droughts. I was heavily involved in the implementation of those action research projects as I was the methodological and technical lead. This initiative enabled me to test and fine tune the method with real life development problems and practitioners in a much wider domain space. The DPPD method, presented in this thesis, provides a tool for development professionals to identify outperformance in different development sectors by mixing analytical insights from traditional and non-traditional data. Such insights should help amplify innovative, locally sourced and evidence-informed solutions to development challenges.
Date of Award31 Dec 2021
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorJulia Handl (Supervisor) & Richard Heeks (Supervisor)


  • International development
  • Positive deviance
  • Non-traditional data
  • Mixed methods
  • Data science
  • High performers
  • Global south

Cite this