Transcript Document

The Time Course of the Oblique Effect in Orientation Sensitivity
Nestor Matthews, Jennifer Cox & Alana Rojewski
Department of Psychology, Denison University, Granville OH 43023 USA
Purpose
Results
It is well established that our ability to see subtle angular
differences is better near cardinal (horizontal or vertical) than
oblique (diagonal) axes. Here, we investigated the whether this
“oblique effect” in orientation sensitivity depends on duration.
Experiment 2: The Effect of Mask Type
Experiment 1: The Effect of Duration
Cardinal
Oblique
Method
Cardinal
1.6
1.4
Oblique
Linear (Oblique)
1
Oblique
y = 0.0029x + 0.2038
R2 = 0.883
0.8
Cardinal
y = 0.0077x + 0.2077
R2 = 0.9792
0.6
0.4
Power (Oblique)
12
Threshold (degrees)
1.2
d'
Gabor patches, and judged whether the second was oriented
‘clockwise’ or ‘anti-clockwise’ to the first. Across trials, we
randomized both the axis (cardinal versus oblique) and the
stimulus duration (8 – 142 mesc). To limit the persistence of
orientation-specific neural responses, circular bulls-eye masks
preceded and followed each Gabor patch.
Exp 2 - We held the stimulus duration constant (108 msec), and
systematically varied whether masks preceded or followed.
Exp 3 - We repeated Exp 1, but now presented the two stimuli
simultaneously on the left and right sides of fixation (see below).
Large Angle
Linear (Cardinal)
Power (Cardinal)
14
Oblique
Exp 1 - Participants foveally viewed sequentially presented
Cardinal
Discussion
10
Oblique
y = 16.193x -0.1488
R2 = 0.8957
8
6
4
0.2
0
0
0
20
40
60
80
100
120
Stimulus duration (ms)
140
160
The data from Exp 2 indicated that for sequential orientation
discrimination, forward masks are more effective than
backward masks, and inner masks are more effective than
outer masks. This is true at both cardinal and oblique axes.
Cardinal
y = 41.717x -0.5468
R2 = 0.9885
2
0
The data from Exp 1 indicated no oblique effect in orientation
discrimination at the briefest stimulus duration (8 msec), where
performance at each axis was poor yet already significantly
better than chance. At all subsequent durations, the oblique
effect was evident, and grew in magnitude with increases in
duration. Indeed, an ANOVA confirmed a significant axis-by
duration interaction, suggesting that the oblique effect in
orientation sensitivity is a dynamic phenomenon.
20 40 60 80 100 120 140 160
Stimulus Duration (ms)
Experiment 3: The Effect of Simultaneity
Data from the two groups (blue & red) overlapped during threshold estimation, showing initial similarity.
During the main experiment, however, performance was significantly worse in the masked group.
Interestingly, the oblique effect under simultaneous stimulation was much reduced.
The data from Exp 3 indicated that across durations the oblique
effect in orientation discrimination is nearly eliminated when
the stimuli are presented simultaneously. This finding may be
difficult to reconcile with purely bottom-up explanations of the
oblique effect.
The Bottom Line
Threshold Estimation
Time
Time
Cardinal
Clockwise
Or
Anti-Clockwise?
Clockwise
Or
Anti-Clockwise?
Oblique
Main Experiment
Precision (d')
Proficiency (d'/RT)
Reaction Time
Physiological studies have suggested that the oblique effect may
occur because fewer visual neurons are tuned to oblique axes
than to cardinal axes. However, if an over-representation of
cardinal neurons alone were sufficient to explain the oblique
effect, one might expect the phenomenon to be constant across
stimulus durations. The present data suggest, instead, that
neural models of the oblique effect in orientation sensitivity
will require consideration of orientation-specific temporal
recruitment.
This poster can be viewed and downloaded at
http://denison.edu/~matthewsn/obliqueeffectcns2004.html