Error detection in reading in English: A validation study of a reading task administered with the help of an eye-tracker

This paper presents a validation study of using eye tracking during a reading task with the aim to investigate reading processes in L2 and to detect errors in reading. I attempted to measure features of reading in L2 in terms of reading time, duration of fixations, and the total number of fixations with the help of the eye tracker. The paper starts with a theoretical overview of eye tracking and error detection in reading, indicating its growing popularity within linguistic circles. The participants of the study were 20 international university students studying in Hungary with their proficiency in English ranging between B2 and C2. The readers were given three texts on a screen to measure their reading performance: the standard text, a text with spelling mistakes, and a text with transpositions. Tobii Pro Lab eye-tracking software was used to conduct the experiment, visualise, and sort the data. As a validation study, this paper mainly uncovers how an eye-tracking instrument can be used for collecting subconscious data and analysing the cognitive processing of adult L2 users while reading, for further investigation of their reading behaviours. As a part of the validation process, it was possible to calibrate the settings of the eye tracking tool for each reader creating more precise data with data loss of a maximum of 6%. This and the findings of the validation study indicated that eye-tracking is instrumental in investigating reading performance. With the help of eye tracking, it was possible to establish that the content of the texts did not have a significant influence on participants' reading performance, nor did the order of the texts. However, there were significant changes in total reading time, total duration of fixations, and the total number of fixations after the mistakes were introduced in the texts.