Abstract:
The purpose of this primarily quantitative study was to compare how young adults
with and without intellectual disabilities examine different types of images. Two
experiments were conducted. The first, a replication and extension of a classic eyetracking
study (Yarbus, 1967), generated eye gaze patterns and data in response to
questions related to the famous painting, The Unexpected Visitor. Both groups exhibited
goal-directed behavior based on the judgment of eight independent raters, an extension to
the original study, but there was a statistically significant difference between the two
groups, based on the judgment of two cooperating raters. Raters could not differentiate
between the scan paths of the young adults with and without intellectual disabilities.
Yarbus’ study was also extended by the inclusion of an interview with the participants.
There was a statistically significant difference in the word count and recollection of
major elements from The Unexpected Visitor between the groups.
The second experiment used eye-tracking technology and a current saliency
model that predicted salient points in images (Walther & Koch, 2006) under two sets of
saliency features. Participants rapidly viewed 30 images of Web sites and other natural
scenes from three different sources. This study found no statistically significant
differences between people with and without intellectual disabilities for teacher created
pictures from a fourth grade geomorphology course and award winning Web sites,
leading to a strong recommendation of a usability rather than accessibility for people with
intellectual disabilities.
Finally, the relative merits of two methods of saliency prediction were compared.
The more recently developed Walther model (Walther & Koch) proved to produce
similar results as the computationally intensive Itti model (Itti, Koch, & Niebur, 1998),
under the conditions of this experiment. This suggests that researchers may use the
simpler model in the future to compare groups.
