User-Centered Assessment of Two Web Browsers: Errors, Perceived Self-Efficacy, and Success

Diane Nohl, Muhammad Pervaiz Meer

University of Hawaii, Honolulu, Hawaii




Internet users (N=21) in two introductory library and information library and information science courses were given a number of routine tasks to perform using Netscape Navigator and Internet Explorer (versions 3) on a Windows 95 operation system. Tasks varied from simple, e.g., going to specified site or adding an address to bookmarks, to more complex, e.g., using a search engine or mailing a web document. Subjects completed a user experience questionnaire that showed their familiarity with Internet software and their self-reported skill level in Web use. Subjects predicted their success with the tasks on perceived self-efficacy rating scales before performing them in the browsers. Frequency and type of errors were analyzed (e.g., navigation errors, scanning errors, formatting errors, and information management errors, among others). After completing each task in a browser, subjects completed perceived difficulty level rating scales. Users in this study had significantly higher success with Explorer, made a variety of errors in several tasks, and those reporting higher self-confidence had significantly higher success scores. The results confirm the expectation developed in prior studies that affective and cognitive measures add valuable user-centered data that can contribute to system design solutions.