An Experimental Comparison of Student Performance in a Web-based Versus a Lecture-based Classroom

Like many of the innovative technologies of the twentieth century, the Internet and its associated applications have been hailed as potentially significant contributors to improving an educational system that has evolved little during the past century and a half. The phenomenal growth of the World Wide Web (WWW) has spurred interest in using it to deliver educational services. In the United States, educational institutions have committed large resources to the use of a Web-based Information System (WIS) to facilitate learning and teaching. However, these commitments are justified primarily on the basis of the non-instructional benefits of the distance model (i.e., increased access, reduction of social bias, efficient utilization of scarce resources such as domain expertise and money, etc.). Little experimental evidence exists to support the premise that instruction using a WIS is comparable to that achieved through traditional methods. This dissertation compared the effectiveness of Web-based andragogy to traditional methods by observing the performance of students learning using both environments. Control and test groups were selected randomly from a population of students enrolled in a computer literacy course. The control group participated in a course delivered traditionally (e.g., through lecture) while the test group participated in a course whose content was delivered using a WIS. The two versions of the course were alike in every respect (e.g., course content, assignments, expected learning outcomes, instructor, methods of assessment, grading criteria, etc.) except for the delivery method used. An examination, consisting of a comprehensive set of questions covering the course\u27s subject material, was used in a pre-test post-test experimental design to determine the increase in learning each student achieved during the course. An empirical measure of student performance was calculated by taking the difference between each student\u27s pretest and post-test examination scores. In addition, students completed separate pre-test and post-test questionnaires, which identified demographic, environmental, and course design characteristics. Correlation and regression analysis techniques were used to discover the nature and strength of relationships between these characteristics and student performance