7.27 Where monitoring data is not feasible or appropriate, bespoke research can be used to collect either process or impact evaluation data. This may be in the form of adding questions to existing surveys5 (which are also useful for providing background information and as a source to sample from or weight back to), if timescales allow, or designing new primary research.
7.28 To meet the requirements of impact evaluations research will need to collect standardised data from both the treatment and control groups to allow for comparison against the counterfactual. Sampling must also be taken into account during the design of both quantitative and qualitative research to ensure that the sample size is large enough to achieve the desired information (for example, statistical power in a quantitative survey) to obtain robust results. These issues are considered in more detail in Chapters 8 and 9 and supplementary guidance.
_________________________________________________________________________
5 Either local or cross-government surveys such as the Labour Force Survey, British Crime Survey, British Social Attitudes Survey or the Family Resources Survey.