Currently, usability tests for web sites with representative users are conducted in laboratory environment that disturbs subjects' natural behavior. These test methods are inefficient for tests that require large number of subjects, because experimenters and subjects should be located in the same place at the same time during the test.
In this study, a remote usability test tool, called 'RIO' has been developed to cope with these laboratory-based experiment problems - the 'synchronism' and the 'unnatural environment'.
The tool is separated into three parts: Project Manager', 'Remote Interaction Observer', and 'Interaction Analyzer'.
'Project Manager' sets up overall experimental parameters and actual tasks to be performed by subjects.
'Remote Interaction Observer', endued with Microsoft Internet Explorer Control, is a modified Web Browser which records user interactions, screen images and elaborate browser events while subject performs given tasks. This module is distributed to subjects as an installable software package. When they finish all the tasks, the captured interaction data is compressed and sent to 'Interaction Data Server' automatically.
'Interaction Analyzer' visualizes interaction data from the 'Interaction Data Server'. It also generates project-scope statistics which facilitate discovering peculiar cases among interaction submissions.
According to the results of 'RIO' experiment with 224 submissions, this tool shows improvements on the time, cost while providing similar quality of data obtainable by laboratory based experiments.