Shortening a survey and using alternative forms of prenotification: Impact on response rate and quality
1 Division of Health Care Policy and Research, Mayo Clinic, 200 First Street Southwest, Rochester Minnesota, 55905, USA
2 Survey Research Center, Department of Health Sciences Research, Mayo Clinic, 200 First Street Southwest, Rochester Minnesota, 55905, USA
3 Division of Digestive Diseases, Hospital Clinico San Carlos, Complutense University, Ciudad Universitaria, Madrid, 28040 Spain
4 Division of Biomedical Statistics and Informatics, Mayo Clinic, 200 First Street Southwest, Rochester, Minnesota, 55905,USA
5 Enteric Neuroscience Program, Division of Gastroenterology and Hepatology, Mayo Clinic, 4500 San Pablo Road, Jacksonville Florida 32224,USA
6 Division of Gastroenterology and Hepatology, Mayo Clinic, 4500 San Pablo Road, Jacksonville Florida 32224,USA
BMC Medical Research Methodology 2010, 10:50 doi:10.1186/1471-2288-10-50Published: 8 June 2010
Evidence suggests that survey response rates are decreasing and that the level of survey response can be influenced by questionnaire length and the use of pre-notification. The goal of the present investigation was determine the effect of questionnaire length and pre-notification type (letter vs. postcard) on measures of survey quality, including response rates, response times (days to return the survey), and item nonresponse.
In July 2008, the authors randomized 900 residents of Olmsted County, Minnesota aged 25-65 years to one of two versions of the Talley Bowel Disease Questionnaire, a survey designed to assess the prevalence of functional gastrointestinal disorders (FGID). One version was two pages long and the other 4 pages. Using a 2 × 2 factorial design, respondents were randomized to survey length and one of two pre-notification types, letter or postcard; 780 residents ultimately received a survey, after excluding those who had moved outside the county or passed away.
Overall, the response rates (RR) did not vary by length of survey (RR = 44.6% for the 2-page survey and 48.4% for the 4-page) or pre-notification type (RR = 46.3% for the letter and 46.8% for the postcard). Differences in response rates by questionnaire length were seen among younger adults who were more likely to respond to the 4-page than the 2-page questionnaire (RR = 39.0% compared to 21.8% for individuals in their 20s and RR = 49.0% compared to 32.3% for those in their 30s). There were no differences across conditions with respect to item non-response or time (days after mailing) to survey response.
This study suggests that the shortest survey does not necessarily provide the best option for increased response rates and survey quality. Pre-notification type (letter or postcard) did not impact response rate suggesting that postcards may be more beneficial due to the lower associated costs of this method of contact.