An Examination of the Lost‐letter Technique1

Abstract
The results of three investigations of the validity and economics of the lost‐letter technique are reported. Although the return rate of the lost‐letter technique was higher than in the case of more obtrusive survey techniques involving hand‐distributed postcard questionnaires and mail surveys, the more obtrusive survey techniques were more economical to carry out, were more accurate in their predictions of election outcomes, and gave results more highly consistent among themselves than with the results obtained by the lost‐letter technique. Furthermore, the results indicated that the act of returning a lost letter may not be strongly affected by the identity of the sender or addressee indicated on the letter.