We propose and illustrate that repeated exposure to stimuli sets increases the size of the saccade amplitudes. Saccadic amplitudes are closely related to the perceptual span and therefore used as a measure for the information intake in an experiment. Studies on expertise have shown that experts have larger saccades than novices and the results on decision times and amount of information search in simple choice experiments indicated that these effects occur also within a single experiment. These findings are particularly important in experiments which use assumptions about perceptual span in order to define areas of interests, focusing their analysis on the information that is ignored in the information search process. If the perceptual span increases over the course of an experiment researchers are at risk of falsely classifying stimuli as not being attended. We conducted a discrete choice experiment in which 68 participants made choices between four alternatives with three different between subject conditions varying in presentation format (verbal matrix, a pictorial matrix, and a realistic product representation). The results consistently demonstrate an increase of the saccade amplitude over the course of the experiment independent of condition. We conclude by discussing our results in the light of the possible increase of the perceptual span and its implications for the research procedure in eye-tracking experiments with a repeated measurement design.