Transactional distance and learner autonomy as predictors of student performance in distance learning courses delivered by three modalities
Description
Growth in distance learning calls for theory-driven research to inform the design and delivery of effective distance instruction (McIsaac & Blocher, 1998). This study used transactional distance and learner autonomy constructs (Moore, 1972; 1983) to explore a predictive model of performance in courses delivered by video-conferencing, telecourse and Internet. Previously, Chen and Willits (1998; 1999) found the underlying factors of these constructs for a path analysis of these factors and students' perceived performance. This study performed between-subjects, survey research in which students self-selected courses delivered by one of the three modalities (Internet, telecourse, compressed video) and sought to answer three research questions: (1) Is the level of either transactional distance or learner autonomy predictive of learning outcomes in distance learning courses overall? Does either interact with delivery modality? (2) Do transactional distance and learner autonomy interact in their effect on learning outcomes in distance learning courses? (3) Overall, are transactional distance and learner autonomy efficacious as the basis for a predictive model of performance for distance learning students? Subjects were 1,324 students in distance learning courses at the University Of New Orleans in fall 2003. Two hundred ninety-two students responded. A 50 item survey instrument measured the predictive constructs transactional distance (a distance of understandings and perceptions that might lead to a communication gap or a psychological space of potential misunderstandings between people), dialog between instructor and student and student and student, structure (rigidity or flexibility of course delivery and organizational aspects), learner autonomy (the learner's ability to select learning objectives, choose methods of study and evaluate progress/achievement), and learning outcomes. Learning outcomes were operationalized as the students' final course grades and students' self-ratings of performance. Multiple regression results showed dialog, structure and learner autonomy to be significant in predictive models of self-ratings of performance and grades, especially in interactions with course type (quantitative, non-quantitative) and prior distance learning experience. Higher structure and dialog tended to have a positive effect on quantitative course grades. Likewise more distance learning experience tended to predict higher grades. Transactional distance was found to predict self-ratings of performance but not grades