Making Mail Surveys More Reliable

Most of the criticism of mail surveys has concerned itself with the problem of biased samples due to small returns (such as o0 per cent to 20 per cent). But with modern methods it is possible to secure from 70 per cent to 93 per cent returns. It is the authors' belief that the whole question of biased samples from surveys made by mail needs to be re-examined in the light of these new techniques in mail sampling. For this purpose we start out with a case history-a recent example of mail surveys which produced a total response of 93 per cent. We show how these results were secured and then examine the findings critically from the standpoint of their reliability.