Search for a command to run...
The awareness of the need for robust impact evaluations in conservation is growing and statistical matching techniques are increasingly being used to assess the impacts of conservation interventions. Used appropriately matching approaches are powerful tools, but they also pose potential pitfalls. We outlined important considerations and best practice when using matching in conservation science. We identified 3 steps in a matching analysis. First, develop a clear theory of change to inform selection of treatment and controls and that accounts for real-world complexities and potential spillover effects. Second, select the appropriate covariates and matching approach. Third, assess the quality of the matching by carrying out a series of checks. The second and third steps can be repeated and should be finalized before outcomes are explored. Future conservation impact evaluations could be improved by increased planning of evaluations alongside the intervention, better integration of qualitative methods, considering spillover effects at larger spatial scales, and more publication of preanalysis plans. Implementing these improvements will require more serious engagement of conservation scientists, practitioners, and funders to mainstream robust impact evaluations into conservation. We hope this article will improve the quality of evaluations and help direct future research to continue to improve the approaches on offer.