The Effect of Interaction Levels on Student Performance: A Comparative Analysis of Web-Mediated versus Traditional Delivery

Article excerpt

This analysis examined the contrasts between online and traditional delivery methods in terms of the impact on student learning and satisfaction for similar course material. It built on the work of other researchers who have asked these types of questions for individual courses and relatively small sample sizes. However, this analysis expanded the number of courses under review and has drawn a larger sample of students than previous studies. The analysis sought to control for student and teacher characteristics, course content, assessment procedures, and so forth, and evaluate student's perceptions of interaction in the two delivery modes. It asked questions about whether varied types of interaction make a difference in a student's performance and satisfaction with the course.

**********

Web-mediated instruction is the fastest growing sector of distance education (U.S. Congressional Web-based Education Commission, 2000). Browser-based delivery has evolved in less than ten years from static HTML pages coupled with email and list-serves to integrated course management suites. Naturally, with the move from the traditional lecture hall to the cyber-classroom, there have been questions regarding instructional effectiveness and efficiency. Some studies of learning outcomes indicate that student performance levels remain about equal but that student perceptions of the course differ (Johnson, Aragon, Shaik, Palma-Rivas, 2000). While these studies have contributed to an overall understanding of the effectiveness of web-mediated instruction, they do little to explore the factors responsible.

One such factor is the level of course interaction. Learning theorist Laurillard (1997) posited that student-teacher interaction is a key component in academic learning. In her "Conversational Framework," she identified four types of activity essential to learning: Discursive, Adaptive, Interactive, and Reflective. Interactive activities refer to the "task/action/feedback" cycle students experience in their own world as they operate on the ideas presented by the instructor. Large class sizes, independent study, and some web-based courses forgo high levels of interaction to achieve instructional efficiency. If Laurillard is right, reduced interaction should result in reduced learning. To date, little empirical research, if any, has been done that correlate distance learning interaction levels with student outcomes. Such research could be an invaluable aid in the design of web-mediated instruction, especially in light of the communication technologies now available in web-ware.

PURPOSE OF THE STUDY

This study investigated course interaction and the impact such interaction may have had on student learning. Specifically, this study addressed the following research questions:

1. When controlling for student and teacher characteristics, course content, and course assessment, is there a general perception that online courses are less interactive than traditional? If so, do students also perceive a difference in the structure or support levels between online and traditional?

2. Does the level of course interaction explain a significant portion of the difference in student performance and student satisfaction? If so, what is the relative importance of each interaction modality (instructor-student, student-student, student-course structure, student support)?

REVIEW OF RELATED LITERATURE

Comparative Analyses of Traditional versus Online Instruction

The education literature is replete with comparative studies of traditional versus technologically mediated instruction. Such studies formed the basis of two meta-analyses that synthesizes research to date Clark (1983) and Liao (1998). These meta-analyses compared conventional lecture/demonstration versus interactive multimedia, computer-aided instruction, or computer simulations. Results were mixed. In his review and meta-analysis of comparative research on the impact of media on learning, Clark concluded, "media does not influence learning under any conditions" (p. …