Bo Chen1, Mark Mitsock1, Jose Coronado1 and Gavriel Salvendy1
1 Hyperion Solutions Corporation, Stamford, CT, U.S.A.
2 School of Industrial Engineering, Purdue University,
West Lafayette, IN, U.S.A.
Conventional usability testing takes place in a usability lab equipped with audio-visual equipment and scanning software. The lab environment allows the experimenter to scan the screen to record the user's interaction, and combined with a think-aloud protocol, to record the users real-time comments and facial expressions. Logging software can gather performance data. The results are high quality data for analysis ( Nielsen 1994).
Remote usability testing through the Internet has the potential to allow the experimenter to gather data of equally high quality without the test subjects having to be physically present at the usability lab. With remote testing, the experimenter at the usability lab uses Internet conferencing software to initiate a collaboration session and to share a software prototype on the usability lab's computer with a subject at a remote site. Because the prototype is running on the usability lab computer, the remote subject's interactions with the software are scanned by the lab's equipment and performance data is logged by the logging software. The remote subjects' real-time comments are recorded by the lab's audio equipment. Optionally facial expression can be recorded if the remote subject has video equipment at their site. Lab-quality data can thus be obtained without incurring prohibitive costs in travel time and expense ( Hartson et al. 1996).
Because many of our clients are multi-national companies with locations all around the globe, we find the potential for gathering high quality usability data over the Intemet very attractive. We wanted to determine whether being at a remote location would have an impact, positive or negative, on the performance of usability test subjects or on their satisfaction ratings of the software. We also