A heuristic-based approach for systematic error correction of gaze data for reading

In eye-tracking research, temporally constant deviations between users’ intended gaze location and location captured by eye-samplers are referred to as systematic error. Systematic errors are frequent and add a lot of noise to the data. It also takes a lot of time and effort to manually correct such disparities. In this paper, we propose and validate a heuristic-based technique to reduce such errors associated with gaze fixations by shifting them to their true locations. This technique is exclusively applicable for reading tasks where the visual objects (characters) are placed on a grid in a sequential manner; which is often the case in psycholinguistic studies.