Reading is a feedback process that requires the integration of different cognitive systems and is an ideal field for exploring the relationships between eye movements with top-down processes. Eye-Tracking (ET) is a non-invasive technology that allows the capture of eye movements with high temporal precision. During the reading process, different characteristics of words, phrases, and the complete story being read influence these movements. Several mathematical models study eye movements during reading short sentences, however, the extension of these findings to natural reading has not been yet studied in depth. The characteristics of the text and the number and variety of the variables involved difficult this modeling task.
Our technique provides a visual representation of the parameters of interest identified in the eye movements data for a typical reading experiment (dwell time, pupil diameter, among others), integrating all of them in a single view.
We focus on short stories, short-length texts that condense a large amount of information. The technique makes it possible to perform the analysis in the context of natural reading and without an a priori formal mathematical model, representing a valuable tool in the first steps of model development to help understanding variables relationships. It allows the analysis of not only short stories reading, but also encourages the generation of new hypotheses to address research activities related to eye movements in reading.