If the success of development efforts to meet the Millennium Development Goals is to be measured, then the impact assessment models employed need to enable the findings to be aggregated and assessed at various levels. Also, if one is to answer the question as to the effects of a project on target groups in different geographical areas, or the effect of different projects on the same target group, the data and information need to be summarised.
Aggregations could possibly be made on a thematic or geographical basis if quantitative methods or scoring-based methods have been used. It is a necessary condition that the same indicators be used. If the indicators have been developed locally with the target group as part of a strong participatory process, then generally binding key indicators must be used in all the studies as a minimum to allow the findings to be aggregated.
There are often challenges in practice when it comes to aggregation. To date, there are only a few approaches, and no fully tried-and-tested methods to analyse the effect of country programmes, sector-wide programmes or programme-oriented community funding.
Another means of summarising data is to produce a meta-evaluation (an evaluation and summary of evaluations) of existing evaluations on a particular subject or a specific region.
The two overviews of impact assessment methods listed below assess various methods in terms of how suitable they are for aggregation (cf. Resources).