The analysis of the results of the referendum presented in the preceding chapter tells us a good deal about variations in voting from place to place. What it cannot do is yield information about the behaviour of individuals or groups of electors. In order to find out how people voted, and to try to explain why they voted as they did, we need to make use of survey data. In this chapter, therefore, we use the results of our postal survey of a random sample of the Scottish electorate, which was undertaken immediately after the referendum (for details see Appendix 2), to explore patterns of turnout and the choices made by voters, attempting to explain who supported which options and why the vote in favour of change was so decisive. We begin, however, by considering voting, as opposed to non-voting, in the referendum.