Tracking attention with the focus-window technique: the information filter must be calibrated.

Abstract

Many researchers have recommended the focus-window tracking (FWT) technique to study the course of information acquisition and visual attention. In FWT, the stimulus is blurred, and the subject scans for information by moving a clear window with a computer mouse. We show that poor calibration of the blur filter will bias search behavior in FWT because diagnostically useful information may leak through a filter even though the subjects are unaware of it. To provide a valid index of information acquisition, FWT requires an information filter that is calibrated to the task.