5.9 To ensure the evaluation provides useful evidence, it is important to consider who the anticipated users of the findings are, and the requirements of policy makers and other stakeholders. These considerations need to be made before the evaluation starts. The findings might be used to:
• support the implementation of the policy;
• inform future decision-making;
• support funding applications;
• improve the ongoing delivery process;
• provide accountability to stakeholders, parliament and the public; and
• contribute to improved knowledge amongst those best able to take advantage of it.
5.10 Thus, when developing the evaluation plan, it is important to understand:
• who the target end-users of the evidence will be. This may include programme managers, policy makers and analysts within the department; other government departments; local authorities and delivery bodies; or key stakeholders including industry bodies, the public, local community groups and other interested parties.
• what are the different expectations for how the results will be used (particularly important for results which may feed back into and affect the ongoing programme delivery) including any expectations on the timing of when the evaluation evidence might feed into decision making;
• what will allow the end users to make most effective use of the evaluation findings. This includes different data requirements, but also presentation of the results, mechanisms for and timing of dissemination. For example, a quantitative cost-benefit assessment of impacts may be required by HM Treasury, while detailed information about effective delivery may be sought by programme managers responsible for the implementation of the same programme on the ground; and
• how robust the evaluation results need to be to support the uses they are intended for, and what level of scrutiny they will be subject to. A decision to support the potential funding and roll-out of a major government initiative is likely to require a high "burden of proof" and hence an evaluation which meets the highest academic standards. Related to this is whether you expect to use average or marginal effects (see Chapter 10, paragraph 10.15 and Table 10.A for further information). An evaluation which is intended to inform specific and limited changes to the way an existing, local intervention is delivered is unlikely to require the same levels of rigour. However, this might limit the generalisability of the evaluation findings and the extent to which they can be seen to add to the evidence base.
5.11 These considerations are therefore likely to influence the evaluation objectives, research questions and evaluation design. By understanding the range of requirements for the evaluation, the questions can be designed to reflect these and methods can be chosen that generate relevant evidence (Step 3).