"Experiments and Experience with Web Lectures and Digital Libraries for Teaching HCI"
Edward Clarkson, Jason Day
College of Computing

12:00 Noon on Thursday, November 16, 2006
ATDC, Centergy Building, Hodges Room 335

Our lab has developed the content and infrastructure for two discipline-specific digital libraries: the Human-Centered Computing Education Digital Library (HCC EDL) and the Visual Analytics Digital Library (VADL). Aside from their use by the HCC and VA communities, we have used them as platforms for DL-related research. This talk will present a system we have developed for search result set visualization in the HCC and VA DLs and discuss our plans for its evaluation and extension.

Naturalistic research has shown that a web lecture intervention that includes multimedia lectures studied before class, short homework assignments, and in-class application activities can increase students' grades and satisfaction. The multimedia lectures, called web lectures, are a combination of video, audio, and PowerPoint streamed over the web. The experimental study briefly outlined in this talk was conducted to understand the contribution of web lectures themselves to the web lecture intervention's success. Educational multimedia design guidelines from Cognitive Load Theory (CLT) and the Cognitive Theory of Multimedia Learning (CTML) were used to evaluate and hypothesize about the learning efficacy of three information-equivalent-Video+Audio+PPT (web lecture), Audio+PPT, PPT+Transcript-and one content-nonequivalent-PPT-Only-educational presentation conditions. 60 randomly assigned participants studied the educational materials and completed a posttest and exit survey. Participants in the web lecture condition performed statistically significantly better on the posttest than all other conditions, and survey responses indicated that participants perceived the combination of modalities used by web lectures as more educationally effective than those used in the other conditions. This study verifies the educational contribution of web lectures to the web lecture intervention, web lectures' educational effectiveness as standalone learning objects, and the value-added of video for educational multimedia. These results were not completely in line with our hypothesis based on CLT and CTML, suggesting these theories' limited applicability for multimedia presentations with characteristics of those used in this study. Several possible factors that might account for the resulting inconsistencies with CLT and CTML are identified, including the visibility of gesture in the video and the length and subject matter of the presentations.