Many eye movement metrics such as the eye fixation duration metric depend, directly or indirectly, on eye fixations. However, eye fixations require an operational definition in order to be correctly classified by the eye movement event detection algorithms. This paper examines the effect of 24 different operational definitions of fixation on the results generated by Identification based on Dispersion Threshold (I-DT), the most popular eye movement event detection algorithm; and hence, the fixation duration metric. The 24 operational definitions are a combination of four different values (0.5, 1.0, 1.3, and 2.0) of the dispersion threshold and six different value (80ms, 100ms, 150ms, 200ms, 300ms, and 400ms) of the duration threshold of the I-DT algorithm. The preliminary results show that there is no statistically significant difference between the various operational definitions of fixation on the fixation duration metric.
Key words: Eye Tracking, Eye Movement Classification Algorithms, Eye Movement Event Detection Algorithms, I-DT, Eye Fixation, Eye Fixation Operational Definition, Fixation Duration
|