Online digital survey methods are used globally as a cost-effective and efficient means of obtaining data about the prevalence and distribution of variables within defined populations and have been increasing in use (Gardner, et al, 2016; Polit & Beck, 2017). The expansion of social media and increased sophistication of digital platforms available for subject recruitment in online survey deployment have increased the options available to researchers engaged in survey research (Wise et al, 2015), but challenges and obstacles remain.
Methods: This study initially sought to conduct a randomized national survey of registered nurse self-efficacy and knowledge for EBP using a descriptive correlation survey design. A power analysis determined 740 subjects were needed. Obstacles obtaining an adequate sample size necessitated modification of the original research plan from stratified random sampling to convenience sampling methods to answer the study questions/hypotheses.
Results: We initially randomized states and hospitals then contacted nurse leaders in these agencies to create our original sampling frame. After approximately 12 months of aggressive subject recruitment including the purchase of state board of nursing rosters of registered nurses and the addition of incentives for online survey completion, we were unable to adequately power the study to answer our question. We therefore sought permission from the Institutional Review Board and STTI to modify the study design to a convenience sampling method at which time we engaged professional networks, hospital-based research councils, posted the call for participation selectively in social media, and invited participation during conferences to obtain the sample size needed.
Conclusion: Eighteen months into our study, and after modifying from stratified random sampling to a convenience sample, we reached our goal of 275 subjects needed to answer our study questions. The use of randomized sampling in online survey deployment has many benefits including rigor and removal of bias associated with convenience sampling. However, the randomized sampling also poses several obstacles, such as reliance on the cooperation of others, variations in databases, surveys landing in Junk mailboxes and time commitment to locating information on websites. Researchers must carefully consider the feasibility of subject recruitment and sampling plan when engaging in online survey research.