Abstract

What is the best way to invite establishments to participate in a web survey? We conducted an experiment by inviting establishment via (i) postal cover letters, (ii) emails or (iii) return envelopes including cards that asked respondents to provide their email addresses in order to invite them via email in a second step. The mode of reminder letters was varied experimentally as well. We analyze the effect of the invitation mode on the participation rate under the control of various establishment characteristics.