Putting the theory of change to work: Process-Outcome Integration with Theory

Enhancing evidence transferability

Calum Davey

London School of Hygiene and Tropical Medicine (LSHTM)

University KwaZulu-Natal
University College London
Columbia University
University British Columbia

January 2020 to December 2022

Secondary data analysis

Uganda, Bangladesh, Nepal, Pakistan

The project seeks to improve methods to better integrate evidence related to the process and outcomes of interventions. It will develop a novel interdisciplinary method, Process-Outcome Integration with Theory (POInT).

1. The study will use case studies to apply POInT to make inferences about the theory of change, describing how interventions work and in which contexts. POInT employs a Bayesian framework to generate a formal, systematic method for assessing the validity of theories of change that can draw on various types of evidence. The case studies will cover disability-inclusive poverty graduation, disability-inclusive youth training and CEDIL-funded evaluations. The outputs will be more accurate and precise estimates of average intervention effects, learning about mechanisms using all of the data and improved generalisable knowledge.

Work with other LSHTM Department for International Development-funded teams for workshops, workshops with implementers and evaluators from universities.

This project will help evaluations glean value from evaluations, and pro-duce evidence that better informs policy. Broader learning will inform evaluation practice in the sector.

• CEDIL Design Paper: POInT Research Design Paper, CEDIL Design Paper 3
• CEDIL Research Project Paper 5: Process Outcome Integration with Theory (POInT): academic report
• Blog post: Turning a theory of change into a Bayesian model : an example from an agricultural intervention
• Blog post: Formalising theories of change as Bayesian causal models and eliciting expert priors over model parameters
• CEDIL conference 2023 presentation by Matt Juden from the project team