Maximising response rates in household telephone surveys
1 Department of Epidemiology and Preventive Medicine, Monash University, The Alfred, Melbourne, Victoria 3004, Australia
2 CRC for Water Quality and Treatment, Salisbury, 5108, South Australia
BMC Medical Research Methodology 2008, 8:71 doi:10.1186/1471-2288-8-71Published: 3 November 2008
Epidemiological and other studies that require participants to respond by completing a questionnaire face the growing threat of non-response. Response rates to household telephone surveys are diminishing because of changes in telecommunications, marketing and culture. Accordingly, updated information is required about the rate of telephone listing in directories and optimal strategies to maximise survey participation.
A total of 3426 households in Sydney, Australia were approached to participate in a computer assisted telephone interview (CATI) regarding their domestic (recycled and/or drinking) water usage. Only randomly selected households in the suburb and postcode of interest with a telephone number listed in the Electronic White Pages (EWP) that matched Australian electoral records were approached.
The CATI response rate for eligible households contacted by telephone was 39%. The rate of matching of electoral and EWP records, a measure of telephone directory coverage, was 55%.
The use of a combination of approaches, such as an advance letter, interviewer training, establishment of researcher credentials, increasing call attempts and targeted call times, remains a good strategy to maximise telephone response rates. However, by way of preparation for future technological changes, reduced telephone number listings and people's increasing resistance to unwanted phone calls, alternatives to telephone surveys, such as internet-based approaches, should be investigated.