10.5061/DRYAD.33KH1
Solman, Grayden J. F.
University of British Columbia
Foulsham, Thomas
University of Essex
Kingstone, Alan
University of British Columbia
Foulsham, Tom
University of Essex
Data from: Eye and head movements are complementary in visual selection
Dryad
dataset
2016
Scene viewing
Visual selection
Head and eye movement
2016-12-21T13:06:14Z
2016-12-21T13:06:14Z
en
https://doi.org/10.1098/rsos.160569
82791 bytes
1
CC0 1.0 Universal (CC0 1.0) Public Domain Dedication
In the natural environment, visual selection is accomplished by a system
of nested effectors, moving the head and body within space and the eyes
within the visual field. However, it is not yet known if the principles of
selection for these different effectors are the same or different. We used
a novel gaze-contingent display in which an asymmetric window of
visibility (a horizontal or vertical slot) was yoked to either head or eye
position. Participants showed highly systematic changes in behaviour,
revealing clear differences in the principles underlying selection by eye
and head. Eye movements were more likely to move in the direction of
visible information—horizontally when viewing with a horizontal slot, and
vertically with a vertical slot. Head movements showed the opposite and
complementary pattern, moving to reveal new information (e.g. vertically
with a horizontal slot and vice versa). These results are consistent with
a nested system in which the head favours exploration of unknown regions,
while the eye exploits what can be seen with finer-scale saccades.
Angular Distribution for Eye and Head MovementsDistribution of gaze shift
angles during scene viewing. Indoor and outdoor scenes were viewed for ten
seconds through gaze-contingent windows controlled by either the eye or
the head. Windows could be Open (no window), Square (equal horizontal and
vertical span), Horizontal (greater horizontal span), or Vertical (greater
vertical span). Gaze shift angles were classified into 16 equal pi/8 bins,
and the data indicate the proportion of shifts in each bin. In addition,
angles were more broadly classified as horizontal (+/- 3pi/16 from the
horizontal axis) or vertical (+/- 3pi/16 from the vertical axis), and the
horizontal bias (H_bias) was calculated as the proportion of horizontal
gaze shifts divided by the sum of horizontal and vertical
shifts.SolmanFoulshamKingstone_DistributionData.xlsx