Effort-Accuracy Trade-Off in Using Knowledge Management Systems

Introduction

Like other information systems, knowledge management systems (KMSs) support the efficient and effective processing of information by facilitating the location of high-quality content from the mass of knowledge they contain (Fang, 2000; Kim & Compton, 2004; Nevo et al., 2003; Orlikowski, 2000). KMSs are shared repositories of potentially useful knowledge to support end users within the same work group or organization (Davenport & Hansen, 1999; Jones & Kochtanek, 2004). KMSs are designed with interfaces that incorporate rating schemes to help users screen out irrelevant, low-quality content (i.e., knowledge). Rating schemes allow KMS users to provide feedback about the quality of content through ratings, potentially improving subsequent content search and evaluation efforts (Shon & Musen, 1999; Standifird, 2001; Wathen & Burkell, 2002). However, future users may be misled if the ratings do not accurately reflect the content quality (Dellarocas, 2003; Resnick et al., 2000). Ratings can be misleading because those supplying the ratings may manipulate ratings intentionally or may rate the content based on a context very different from the users’ current context (Cosley et al., 2003; Cramton, 2001). Consequently, users relying on misleading ratings may select high-rated, low-quality content that is obsolete and incomplete to use in their particular task (Cosley et al., 2003; Melnik & Alm, 2002).