What is it?

The DPPD method leverages different types of digital data to identify and characterise outperformers in complex challenges.

It builds on the Positive Deviance approach, which is based on the observation that in every population there are individuals or communities who achieve better results than their peers despite facing similar challenges and limitations.

The DPPD method was first described by Basma Albanna and Richard Heeks and then further developed through its application in a global initiative by the GIZ Data Lab, Pulse Lab Jakarta, the UNDP Accelerator Labs Network, and the University of Manchester.

Why use it?

The DPPD method helps understand what works and how. These insights could support the scaling of locally-sourced innovative solutions.

Bottom-up approach

The method capitalizes on the resources, experiences and knowledge of a community as the ultimate starting point in the search for solutions. These solutions are tailored to the local context, making them more sustainable and less vulnerable to social rejection.

Leverage non-traditional data

The abundance of digital data relating to performance outcomes of individuals, communities, and spaces create new opportunities for social innovation. We leverage such readily-available data to identify and learn from outliers.

Combine big and thick data

The DPPD method provides a new way to integrate insights from both big and thick data. Big data is used to identify who is outperforming, while thick data is used to explain why. This increases our ability to extract value from big data while avoiding potential context loss.

Evidence-informed solutions

Digital data, often covering wide geographic areas, yields a large sample of positive deviants, strengthening our ability to make inferences about their outperformance. This provides a tool for the design of effective interventions, backed by statistical evidence.

How does it work?

The DPPD method enables us to integrate non-traditional data in the Positive Deviance approach:

1

Assess Problem-Method Fit

Define the problem and scope of the intervention. Check if DPPD is a suitable and feasible method by assessing the required data and capabilities and by ensuring that potential outcomes are desirable for the community.  


2

Discover Factors Underlying Outperformance

Collect field data to identify distinguishing factors between positive and non-positive deviants. Analyze the results to uncover practices and strategies that explain the success of positive deviants.


Determine Positive Deviants 

Divide the study population into groups and measure their performance to identify potential positive deviants. Conclude this stage with a preliminary validation of the identified potential positive deviants.

3


4

Design and Implement Interventions 

Assess the potential of the identified solutions for replication and scaling. Use the insights to design and implement community and policy interventions together with the community.


Monitor and Evaluate Interventions

Measure, monitor, and evaluate the effectiveness of interventions that aim to scale and amplify the uncommon practices, strategies, and other factors underlying the performance of positive deviants.

5


Learn more about how it works from our Handbook

Resources

Take a look at these publications to learn more about the method.

Positive deviance, big data and development: a systematic literature review

By Basma Albanna and Richard Heeks

Read the paper

Data-powered positive deviance: combining traditional and non-traditional data to identify and characterize development-related outperformers

By Basma Albanna, Richard Heeks, Andreas Pawelke, Jeremy Boy, Julia Handl, and Andreas Glücker

Read the paper

Learning from the edges: lessons learned from applying the data-powered positive deviance method to identify grassroots solutions using digital data

By Andreas Pawelke, Andreas Glücker, Basma Albanna, and Jeremy Boy

Read the report

Leveraging the positive deviance approach using big data

By Basma Albanna

Check out the PhD thesis

The data-powered positive deviance handbook

By Basma Albanna and Andreas Pawelke

Check out the Handbook