|
Title
|
|
|
|
No mechanism without context : strengthening the analysis of context in realist evaluations using causal loop diagramming
| |
|
Author
|
|
|
|
| |
|
Abstract
|
|
|
|
| Realist evaluation is an approach with a strong emphasis on causal mechanisms and the context in which they are triggered. However, recent reviews of published realist evaluations show that context is often understudied. This is problematic, as a thorough understanding of the relationship between context and causal mechanisms is crucial in assisting policymakers to make appropriate and targeted decisions that improve the intervention. Therefore, we set out to test whether combining realist evaluation with the “systems thinking” approach and, more specifically, causal loop diagramming, could help strengthen the analysis of context. We did this through a study of a performance‐based financing (PBF) intervention in the Ugandan health care sector by the Belgian development agency, Enabel. PBF allocates funds to health workers and/or health facilities based on their performance, and introduces additional management support tools, provides extra monitoring and supervision, and promotes community participation in management issues, among other activities. In this case, we found that the proposed combined methodological approach indeed adds value to the analysis, as it leads to insights into the role played by the underlying system that otherwise may have been overlooked. Moreover, such information may provide clear directions to policymakers on how to improve the intervention in a sustainable way. Finally, causal loop diagrams help to visualize complex causal interactions and to communicate them to policymakers. |
| |
|
Language
|
|
|
|
English
| |
|
Source (journal)
|
|
|
|
New directions for evaluation / American Evaluation Association. - San Francisco, Calif.
| |
|
Publication
|
|
|
|
San Francisco, Calif.
:
2020
| |
|
ISSN
|
|
|
|
1097-6736
| |
|
DOI
|
|
|
|
10.1002/EV.20424
| |
|
Volume/pages
|
|
|
|
167
(2020)
, p. 101-114
| |
|
Full text (Publisher's DOI)
|
|
|
|
| |
|
Full text (open access)
|
|
|
|
| |
|