The strange case of online surveys: response issues and respondent characteristics

This research contributes to existing knowledge about collecting data online by analysing multiple data sets on key criteria including overall representativeness of the samples, response and break-off rates, timeliness of response and reminder effects. Across eight online surveys that initially utilise e-mail lists we found an overall response rate of 17% (n=3,687), where 60% of respondents partake within 48 hours of receiving an e-mail invitation. A reminder, generally sent between three to five days later, resulted in another 30% of responses. In general, respondents who are younger (18-30 years of age), single or in ‘other’ types of family relationships, and belong to a low-income bracket (annual household income < $30,000) are likely to belong to the group of Laggards (responding after 73+ hours). It is clear, that a successful study can be run in less than a week, sending out a reminder 48-72 hours after the initial invitation and closing the survey one to two days later; based on our results this time-wise approach still captures 90% of respondents. This study must be viewed in light of some key limitations. Firstly, we have not considered whether there are differences in responses on the preference measures for Leaders, Early or Late Majority or Laggards and secondly, we have only considered the demographic characteristics of responders. The information can be utilised by researchers in the planning and management process of online surveys, especially since the results clearly indicate that an online survey can successfully be run over a maximum of five days, and as we all know: time is money.