Abstract
Motter and Belky (1997) analyzed monkey eye movements during search tasks. They took the relative directional headings for consecutive saccades and found a slight directional bias against saccades to areas between the previously fixated stimulus and the current fixation location. In the current research, an analysis of human eye movements during free viewing was made. Eight images of natural scene were tested with 118 subjects. The subject viewed every image freely for 10 sec. The relative directional headings for consecutive saccade were broken out of the data set and analyzed for directional biases. Saccade direction polar histograms average across subjects showed directional biases: a consecutive saccade took a straight line slightly more than a left or right turn, and it went backward definitely more than the other directions.