The classic adage is that the selection of a research method depends on the research question that is the focus of attention. Thus, if you are trying to find out what the result of a national election might be in terms of voting preferences, you would need a method that could deliver a representative sample of voters and ask about their intentions in order to have the best chance of finding an answer. So, the method chosen would be an opinion poll, although, as the failure of polls to correctly predict the outcome of some elections (notably, the 2015 general election in the UK) suggests, the chosen method might not always deliver a successful outcome. Yet, the advice that you should let your question determine your choice of method remains generally good counsel; however, in the world we are interested in – connecting social science to public policy – we would argue that the selection of a method requires the consideration of a range of other issues. Above all, the issue of choosing a method depends not only on your question at stake, but also on the state of play you face in policy making.

In particular, it is important to recognise at the very least that the character of policy making is influenced by time constraints, by the preferences or needs of relevant interests, and by the limits of available resources to deal with an issue. These are all points raised in Chapter One and they explain, of course, why evidence is not the sole criterion for making policy choices. In addition, there are times when policy options are more open and times when they are more closed. The policy context, as a result, directs the choice of research method that might be appropriate. The general argument is well presented by Bardach and Patashnik (2016: 14), in their classic, if not unproblematic, book providing advice to policymakers, which is in its fifth edition as of 2016: ‘Thinking and collecting data are complementary activities: you can be a much more efficient collector of data if you think, and keep on thinking, about what you do and don't need (or want) to know, and why’.