NGO support toolkit (English)  >  Monitoring and evaluation  >  Evaluations 

Outcome/impact evaluations

Outcome evaluations tend to focus on immediate outcomes of activities, while impact evaluations tend to focus on longer-term change resulting from activities. In practice, end-of-project evaluations tend to include both types of questions.

Reliable measurement of programme impact is often beyond the scope of even large NGOs. One reason is that impact evaluations generally require baseline information. Unfortunately, many organisations tend not to think about evaluation until they are nearing the end of a project, and by then it is too late to collect baseline data. In this situation, one option is to undertake a ‘retrospective’ impact evaluation, which seeks to build up a picture of the baseline through; for instance, interviews with those involved from the outset. This ‘re-created’ baseline is clearly much less reliable than data collected at the project outset because individual memories tend to be affected by the programme itself. Collection of baseline data, for instance on knowledge, attitudes and practices regarding HIV/AIDS, can be expensive and time consuming. It may be possible instead, to use ongoing surveillance data collected by the government, or other larger organisations.

Impact evaluations are also challenging because of the difficulty of attributing the observed changes to the work of the NGO (as opposed to other local and national activities being undertaken in the area). Confidence that the observed change is due specifically to the NSP, can only come from experimental evaluation designs (for instance, a randomised design where groups or individuals are randomly assigned to receive either the intervention or act as a control). However such designs can raise ethical concerns and are neither affordable nor practical for many NGOs. A practical compromise is to undertake a ‘before and after’ evaluation, whilst carefully tracking other activities in the programme area that may have brought about change (whether positive or negative). Consistent tracking allows an NSP to attribute change with a reasonable degree of confidence. However, it is important to be realistic about what can be demonstrated through an impact evaluation.

• Send to a friend
• Printer-friendly