Assumptions in theories of change.

This paper argues that assumptions in a theory of change are the causal connections, events, and conditions that need to be realized for the intervention to work. Using an example of an intervention aimed at improving educational outcomes for girls in a conservative region, two kinds of assumptions are discussed: cause-effect assumptions and causal-link assumptions. Implications for the use of theories of change, including their use in setting causality and the utility of evidence in argument for learning about and testing assumptions are also discussed. The need for an explicit description of what is meant by the term 'assumptions' in association with a theory of change is also highlighted.