Web-Based Versus Traditional Paper Questionnaires: A Mixed-Mode Survey With a Nordic Perspective

• Published in: Journal of Medical Internet Research
• Main Authors: Hohwü, Lena, Lyshol, Heidi, Gissler, Mika, Jonsson, Stefan Hrafn, Petzold, Max, Obel, Carsten
• Format: Article in Journal/Newspaper
• Language: English
• Published: JMIR Publications 2013
• Subjects: Computer applications to medicine. Medical informatics; R858-859.7; Public aspects of medicine; RA1-1270; Norway; Iceland
• Online Access: https://doi.org/10.2196/jmir.2595; https://doaj.org/article/8b56ce50c4684c829499a1d6eb1018dc

BackgroundSurvey response rates have been declining over the past decade. The more widespread use of the Internet and Web-based technologies among potential health survey participants suggests that Web-based questionnaires may be an alternative to paper questionnaires in future epidemiological studies. ObjectiveTo compare response rates in a population of parents by using 4 different modes of data collection for a questionnaire survey of which 1 involved a nonmonetary incentive. MethodsA random sample of 3148 parents of Danish children aged 2-17 years were invited to participate in the Danish part of the NordChild 2011 survey on their children’s health and welfare. NordChild was conducted in 1984 and 1996 in collaboration with Finland, Iceland, Norway, and Sweden using mailed paper questionnaires only. In 2011, all countries used conventional paper versions only except Denmark where the parents were randomized into 4 groups: (1) 789 received a paper questionnaire only (paper), (2) 786 received the paper questionnaire and a log-in code to the Web-based questionnaire (paper/Web), (3) 787 received a log-in code to the Web-based questionnaire (Web), and (4) 786 received log-in details to the Web-based questionnaire and were given an incentive consisting of a chance to win a tablet computer (Web/tablet). In connection with the first reminder, the nonresponders in the paper, paper/Web, and Web groups were also present with the opportunity to win a tablet computer as a means of motivation. Descriptive analysis was performed using chi-square tests. Odds ratios were used to estimate differences in response rates between the 4 modes. ResultsIn 2011, 1704 of 3148 (54.13%) respondents answered the Danish questionnaire. The highest response rate was with the paper mode (n=443, 56.2%). The other groups had similar response rates: paper/Web (n=422, 53.7%), Web (n=420, 53.4%), and Web/tablet (n=419, 53.3%) modes. Compared to the paper mode, the odds for response rate in the paper/Web decreased by 9% (OR 0.91, 95% CI 0.74-1.10) ...