Response rate, incentives and timing of online surveys: A study of agriculture researchers in Kenya

Abstract

With the increase in internet connectivity, web-based surveys will increasingly be used as a tool for data collection characterized by low financial resource implication, quick response time, and no need for separate data entry. This study had the objective of learning how different survey window lengths and incentives affect responses to online surveys. The study was carried out in the context of an online conjoint survey involving the ranking of different domestic biogas plant attributes presented to 345 respondents from the Kenya Agricultural and Livestock Research Organization (KALRO). Email invitations to the survey were sent in three waves spaced one week apart. The respondents were assigned to 4 blocks composed of an incentive in the form of a lottery of either (zero, $50, $100, or $150) for having completed the survey. Within these groups, a timing (one, two, or three weeks) by which the survey would be opened to receive responses was included as an additional treatment. With a response rate of 11.8 percent, neither the incentive nor the time to respond had an impact (p<0.05) on the decision to respond. Sixty-five percent of those responses were received within the window indicated in the email invitation. The distribution- free Kruskal-Wallis H test revealed no jointly significant association between the response speed and the incentive nor the time granted treatments. Similar results are obtained through the non-parametric equivalent of the ANOVA (Jonckheere-Terpstra trend test) which reveals that neither of the treatments separately or jointly influenced response speed. However, longer deadlines may induce some response. Results also show an effect suspected to be that of a reminder.