This was the ﬁrst method of studying happiness to be used, and a large number of surveys have now been carried out. Some of these were on a very large scale, with 160,000 respondents in one case; this means that some good statistical analyses can be made. At ﬁrst the interest was mainly in the percentage of individuals who were happy or unhappy. Then there were many surveys investigating the statistical relationship of self-reported happiness and other variables, such as age, being married, being employed and so on; Campbell et al. (1976) was an inﬂuential early example. Then it was realised that some of these variables were correlated with one another so that one might really reﬂect the inﬂuence of another. Multiple regressions became popular with sociologists to show the independent eﬀect of diﬀerent variables. We referred earlier to the 630 studies, many using these methods, which were re-worked by Veenhoven (1994). We learn for example that race has little eﬀect on SWB if income, education and similar variables are taken into account.