Using eye tracking to address limitations in think-aloud protocol

This paper presents preliminary findings from an eyetracking study that investigates the validity of responses provided by participants who have been instructed to use think-aloud protocol during a Website usability study. Ten participants were given four information-finding scenarios to perform on the Washington State Department of Licensing website and they were instructed to think aloud. While performing these tasks, an eyetracking system determined where on the screen the subjects looked. Videotapes of the sessions containing the participant's computer screen, mouse and eye movements, and verbalizations were analyzed to determine the degree validity offered think-aloud protocol. Initial findings suggest that some subtle cues indicating participants' expectations of information and link location, and their confidence in link names may be lost when using think-aloud protocol and observation alone.