Estimating Gaze Duration Error from Eye Tracking Data

What is it about?

When digital eye tracking applications are used they generate a sequence of fixation points on the screen. These fixation points will contain some error drawn from a distribution that often depends on the true position of fixation within the bounds of the screen. When we use these fixation points to generate gaze duration estimates, as is done in area-of-interest studies, then we are at the mercy of the fixation error. The relationship between fixation error and gaze duration error is not directly derivable, but in this work we develop a tool for using simulations to estimate the distribution in gaze duration error from an underlying fixation error profile.

Featured Image

Photo by Harry Quan on Unsplash

Photo by Harry Quan on Unsplash

Why is it important?

Any study that relies on fixation point eye tracking to do an area-of-interest study needs to have some idea of the expected error in gaze duration estimates. This research presents an open source software package that can be used to produce these estimates.