Social Media, Sociality, and Survey Research

CHAPTER 5

Virtual Cognitive Interviewing Using Skype and Second Life

Elizabeth Dean, Brian Head, and Jodi Swicegood,

RTI International

Surveys are prone to numerous sources of error. Survey methodologists refer to the combination of these errors as total survey error (TSE) (Biemer & Lyberg, 2003). Measurement or response error is a major component of TSE and stems from the question-and-answer process (FCSM, 2001; Groves et al., 2004; Groves & Lyberg, 2010). The primary purpose of cognitive interviewing is to identify and minimize measurement error (Willis, 2005; Beatty & Willis, 2007). Cognitive interviews have been used to help identify question comprehension problems (Harris-Kojetin et al., 1999; Beatty, 2002), vague or inconsistent instructions (Carbone, Cambell, & Honess-Moreale, 2002; Zuckerberg & Lee, 1997); problems with logic or flow (Redline et al., 1998; Harris-Kojetin et al., 1999); and issues with participant burden or questionnaire complexity (Beatty, 2002; Miller, 2003).

Over the past 30 years, cognitive interviewing has become one of the most commonly used questionnaire pretesting techniques among academic, government, and private survey research organizations. For much of its history, cognitive interviewing has been conducted in laboratories housed at these institutions. Advantages to hosting interviews in a laboratory include control over the interviewing environment, the capability to use technologies that require proximity to the laboratory (e.g., video recording ...