7.1 Whatever evaluation approach is used, data will be required to evaluate a policy; what data will depend on the types of evaluation proposed and the research questions to be answered. There are four main types of data which, if planned for, might be able to play a key role in supporting both process and impact evaluations:
• existing administrative data that has not been collected specifically for the evaluation;
• long term, large scale, often longitudinal, structural survey data managed by central governments or the Office for National Statistics;1
• monitoring data or performance management data that are already being collected to support the administration of the policy; and
• new data collection needed to support the evaluations information needs.
7.2 The availability of administrative or general long-term scale structural survey data should always be considered at the design stage of an evaluation because they have the potential to be important sources of background or explanatory data, for example unemployment rates used to explain crime trends.
7.3 In certain cases, where the evaluation has a sufficiently long lead in time, it might be possible to influence the collection of certain information through these sources, but this should not be relied on as a way to provide detailed project specific information.
7.4 As any administrative and long-term survey data will, by their nature, be being collected anyway this chapter will focus on monitoring data (which in some cases will be a sub-set of general administrative data relevant to the operation of the policy or programme), new data collection and data collection tools, before ending with a discussion about ethical and data protection considerations.
_________________________________________________________________________
1 For example the Labour Force Survey, the British Crime Survey, the Wealth and Assets survey, the English Longitudinal Study of Ageing or the Birth Cohort Studies.