To the Editor:
Nonresponse in population surveys is an increasing concern in health and social sciences. The magnitude of bias associated with nonresponse cannot be easily quantified or adjusted for in the estimation of prevalence and other parameters.1,2 By using various modalities, Dillman et al3 propose an evidence-based method to maximize response rates to surveys using various modalities. Sending an invitation or prenotice letter by the use of registered post or recorded delivery is cited as effective in increasing response rates for postal surveys.4
We used a modified version of the Dillman et al3 method to recruit parents of adolescents to complete a Web survey about their attitudes toward adolescent alcohol consumption and their supply of alcohol to adolescents. We randomized the sample to a prenotice letter sent by registered post versus standard mail, with the expectation that registered post would yield a higher response rate.
We performed the trial at a high school in the Hunter region of New South Wales, Australia. We sent a prenotice letter to all parents of eligible children at the school, describing the purpose of the research, and indicating that they would soon receive an email invitation, and that completing the survey would give them entry to a draw to win an iPad. The letter enclosed a pen as a token of appreciation for the recipient’s time.
On day 1, we randomly allocated 472 addresses to registered post (n = 226) or standard mail (n = 245). On day 9, a message was sent to the email address corresponding to each postal address. On days 16 and 30, reminder emails were sent to those who had not responded, and a final reminder was sent on day 37. Registered post cost AUD 5.70 per letter in comparison to AUD 0.45 for standard mail.
Seven of the registered post letters were returned as undeliverable, bringing the number of registered post letters delivered to 219. In Australia, if the registered post letter cannot be delivered in person, a notice is placed in the letterbox requesting that that the addressee collect it from the post office during business hours. One recipient of such a notice complained about this inconvenience. Undeliverable mail or incorrect addresses cannot be recorded for standard mail.
In the registered post group, two parents registered their ineligibility (child outside age range) and 33% (73/219) completed the survey. In the standard mail group, one recipient registered ineligibility and 36% of parents (88/245) completed the survey (difference = 3% [95% confidence interval = −6% to 11%). These results indicate no advantage of registered post for the delivery of a prenotice letter for a Web survey.
School of Medicine and Public Health
and Priority Research Centre for
Hunter Medical Research Institute
University of Newcastle
1. Caetano R. Non-response in alcohol and drug surveys: a research topic in need of further attention. Addiction. 2001;96:1541–1545
2. Meiklejohn J, Connor J, Kypri K. The effect of low survey response rates on estimates of alcohol consumption in a general population survey. PLoS One. 2012;7:e35527
3. Dillman D, Smyth J, Christian L Internet, Mail, and Mixed Mode Surveys. The Tailored Design Method. 20093rd ed Hoboken, N.J. John Wiley and Sons
4. Edwards PJ, Roberts I, Clarke MJ, et al. Methods to increase response to postal and electronic questionnaires. Cochrane Database Syst Rev. 2009(3):MR000008