Every research methodology for data collection has both strengths and limitations, and this is certainly true for transaction log analysis. Therefore, researchers often need to use other data collection methods with transaction logs. In this chapter, we discuss surveys as a viable alternate method for transaction log analysis and then present a brief review of survey research literature, with a focus on the use of surveys for Web-related research. The chapter then identifies the steps in implementing survey research and designing a survey instrument. We conclude with a case study of a large electronic survey to illustrate what surveys in conjunction with transaction logs can bring to a research study.
Review Of Literature
Although surveys have been used for hundreds of years, the Web provides a remarkable channel for the use of surveys to conduct data collection (Jansen, Corley, & Jansen, 2006). Many of these Internet surveys have focused on demographical aspects of Web use over time (Kehoe & Pitkow, 1996) or one particular Website feature (Waite & Harrison, 2002). Treiblmaier (2007) presents an extensive review of the use of surveys for Website analysis.
Survey respondents may include general Web users or samples from specific population. For example, Jeong, Oh, and Gregoire (2003) surveyed travel and hotel shoppers. Huang (2003) surveyed users of continuing education programs, and Kim and Stoel (2004) surveyed female shoppers who had purchased apparel online.
For academic researchers, a convenience sample of students is often used to facilitate survey studies, including the users of Web search engines (Spink, Bateman, & Jansen, 1999). McKinney Yoon and Zahedi (2002) used both undergraduate and graduate students as their sample examining use of a Website. The major advantages of using students that are often cited include a homogeneous sample, access (Huizingh, 2002), their familiarity with the Internet (Jansen & McNeese, 2005), and creation of experimental settings (Rose, Meuter, & Curran, 2005). There are concerns in generalizing these results (Abdinnour-Helm, Chaparro, & Farmer, 2005), most notably for Websites and services where students have limited domain or system knowledge (Kim & Stoel, 2004; Koufaris, 2002). However, as a sample of demographic slice of the Web population, students appear to be a workable convenience sample with results from studies with students (c.f., Jansen & McNeese, 2005; Kellar, Watters, & Shepherd, 2007) similar to those using other sampling methods (c.f., Hargittai, 2002; Kehoe & Pitkow, 1996).