Many problems with evaluations can be traced back to the Terms of Reference (ToR) - the statement of what is required in an evaluation. Many ToRs are too vague, too ambitious, inaccurate or not appropriate.
One of the ways of improving Terms of Reference is to conduct an evaluability assessment, and to use this to inform the ToR
There are various forms of Evaluability Assessment but they all focus on assessing whether the project/program and the organisation are ready to be evaluated.
This can involve:
- clarifying what the program is trying to achieve and how (that is, articulating its theory of change, or developing a logic model),
- reviewing available data to see if this is adequate to support an evaluation or whether it suggests there is sufficient information to avoid or delay an evaluation (for example if it is already clear that implementation has been weak, there might not be much point in doing an evaluation of its effectiveness)
- checking whether there is genuine commitment to using an evaluation within the organisation
In terms of the BetterEvaluation Rainbow Framework, Evaluability Assessment addresses elements of DEFINING (developing an initial description, developing a logic model), MANAGING (Understand and engage stakeholders), and DESCRIBING (Reviewing existing records and data).
Doing an Evaluability Assessment can determine whether or not the door is truly open for an evaluation.
Check out the page about Evaluability Assessment in BetterEvaluation, including links to selected examples and guides.
For a comprehensive list of resources, see Rick Davies' recent (Nov 2012) comprehensive review of resources.
Rick Davies: A Bibliography on Evaluability Assessment
Have you used Evaluability Assessment? Do you use other processes to ensure there is a good understanding of what is to be evaluated and why before beginning an evaluation?