A randomised trial found online questionnaires supplemented by postal reminders generated a cost-effective and generalisable sample, but don't forget the reminders.

OBJECTIVE: To compare the response rates, data completeness and representativeness of survey data produced by online and postal surveys. STUDY DESIGN AND SETTING: A randomised trial nested within a cohort study in Yorkshire, United Kingdom. Participants were randomised to receive either an electronic (online) survey questionnaire with paper reminder (N=2982), or paper questionnaire with electronic reminder (N=2855). RESULTS: Response rates were similar for electronic contact and postal contacts (50.9% versus 49.7%, difference = 1.2%, 95% confidence interval -1.3% to 3.8%). The characteristics of those responding to the two groups were similar. Participants nevertheless demonstrated an overwhelming preference for postal questionnaires, with the majority responding by post in both groups. CONCLUSION: Online survey questionnaire systems need to be supplemented with a postal reminder to achieve acceptable uptake, but doing so provides a similar response rate and case mix when compared to postal questionnaires alone. For large surveys, online survey systems may be cost saving.