Session S2C 978-1-4244-6262-9/10/$26.00 ©2010 IEEE October 27 - 30, 2010, Washington, DC 40 th ASEE/IEEE Frontiers in Education Conference S2C-1 Work in Progress - Design and Assessment of an Asynchronous On-Line Graduate Statistics Course Larry G Richards University of Virginia, lgr@virginia.edu Abstract – Since 1985, the University of Virginia has offered a graduate course Probability and Statistics for Engineers and Scientists. A similar course has been offered at Virginia Commonwealth University. Both are essential components in Virginia’s Commonwealth Graduate Engineering Program – a distance learning initiative involving universities, industry and government partners. This course has two distinct audiences - full time students in-residence at UVA, and those who attend classes through CGEP while maintaining full-time jobs. In this paper, we review the results of the first offering of this course in an asynchronous on-line mode. All lectures were recorded in Camtasia and coordinated with PowerPoint slides (available as PDFs). Student reactions were assessed through surveys and anonymous feedback, as well as comments and e-mails. The level of acceptance of the on-line lectures was surprising. Most students reported high levels of satisfaction with the quality of the material, mode of delivery, flexibility and convenience. Index Terms – Asynchronous learning, Distance Education, On-line Learning, Electronic Instructional Media. INTRODUCTION Probability and Statistics for Engineers and Scientists has been offered at The University of Virginia and Virginia Commonwealth University for over 25 years in both traditional and distance learning formats [4, 5]. As enrollments grew, so did a series of logistical problems. In large cities, students have difficulty commuting to the receive sites. And because most off-grounds students work full-time, business and military assignments frequently result in irregular class attendance. When we started recording classes and posting streaming video on the Internet, class attendance dropped off, but student performance did not. Many students reported accessing class materials from airports, hotel rooms, homes and offices. Even on-grounds students often miss classes as they pursue their research and attend conferences and meetings. The different expectations, learning styles, and study habits of these two distinct groups caused some concern in the past, so we decided to assess their reactions to and performance in a new mode of course delivery. For several reasons we had decided to offer this course on-line with asynchronous presentation of all materials. Traditionally this class involved thirty-two 75-minute class periods where lectures were presented. A standard textbook [1] and software package [2, 3] were required, and homework, exams, and assignments were used to assess student performance. We have incorporated cooperative learning activities into these classes [6, 7] to try to overcome competitive pressures and encourage students to help each other master the material. In the new version of the course, the lecture material has been segmented into a series of 30 minute modules (chunks). The modules are clustered into Units. Each unit focuses on a coherent set of topics – data analysis and display, basic probability and random variables, estimation, hypothesis testing, analysis of variance, correlation and regression, and quality control and assessment. The first four units are required of all students; the last three are advanced topics, and each student is free to choose which one(s) to pursue. Students have access to the various units during particular time frames, but they decide when and where to study them. Each module has embedded assignments to facilitate active learning. Tests and quizzes are taken in defined periods to insure steady progress by the students. Learning is self- directed, but paced. Supplementary modules review prerequisite knowledge, and others provide deeper coverage of selected concepts and methods. The goal is a personalized path through the material relevant to the students’ interests and needs. The design of the course was influenced by similar efforts spanning over 60 years. Don Dulany, Jerry Uhl, and Burks Oakley at Illinois developed unique delivery systems for courses in psychology, calculus, and electric circuits, respectively. At Michigan State, Peter Signell constructed a self paced introduction to Physics based on short modules focused on key topics, and Ray Frankmann taught Statistics by analyzing the relations between concepts and procedures and revealing these cognitive maps to his students. The organizational and logistic aspects of our course are based on well established “best practices”, and are grounded in the literatures of the learning sciences and educational theory. PILOT ON-LINE STATISTICS COURSE This semester we had 63 students in the graduate statistics class - 20 on-grounds at the University of Virginia and the