Sensation and Perception Presentation

1
Sensation and Perception Presentation

• Human Orientation Discrimination Tested with
  Long Stimuli
• Guy A. Orban, Erik Vandenbussche, and Rufin
  Vogels
• 1984

2
Question

• Are S cells of area 17 (V1) involved in
  orientation discrimination?

3
Alternatives

• YES, S cells are involved in orientation
  discrimination.
• NO, S cells are not involved in orientation
  discrimination.

4
Logic
Prediction 1

• Long stimuli should have better discrimination
  for orientations within 15 from a principal
  meridian. Equally worse outside of the 15
  range.
• Why? Because the there are more S cells tuned
  for those orientations than for the others.

5
Logic
Prediction 2

• Because of the length summation of S cells, the
  meridional variations should be much stronger for
  long lines than for short lines.

6
Methods

• Stimuli
• Subjects were presented with an oriented
  fluorescent rod inside an otherwise dark box at a
  viewing distance of the standard 57 cm.
• Stimuli were presented briefly and successively
  with a short gap between them.
• Subjects had to respond within an allotted time.
  
• Subjects
• Some male and female students between 20 and 25
  years of age. Some of the subjects were naïve
  about the experiment and some were not.
• They all had normal or corrected vision.

7
Methods

• Method of Constant Stimuli
• Used to determine the JND in orientation at a
  given reference orientation.
• Subjects were supposed to press a button to
  indicate whether the stimulus was clockwise or
  counterclockwise from the reference orientation.
• Each orientation was presented randomly 64
  times. JND was based on 320 responses.

8
Results

• JNDs in orientation are much smaller at
  principal orientations (horizontal or vertical)
  than at oblique orientations. (Fig. 1)
• (Replication)
• JND in orientation increases as a function of
  obliquity from the principal orientation up to
  20 obliquity and then levels off. (Fig. 3)
• (Prediction 1)
• Long stimuli give a strong oblique effect in
  orientation acuity whereas short stimuli give a
  weaker oblique effect. In other words, the
  oblique effect increases as line length
  increases. (Fig. 4)
• (Prediction 2)

9
Figure 1

• Distribution of JNDs in orientation at the four
  main reference orientations (horizontal,
  vertical, left oblique, right oblique).

10
Results

• JNDs in orientation are much smaller at
  principal orientations (horizontal or vertical)
  than at oblique orientations. (Fig. 1)
• (Replication)
• JND in orientation increases as a function of
  obliquity from the principal orientation up to
  20 obliquity and then levels off. (Fig. 3)
• (Prediction 1)
• Long stimuli give a strong oblique effect in
  orientation acuity whereas short stimuli give a
  weaker oblique effect. In other words, the
  oblique effect increases as line length
  increases. (Fig. 4)
• (Prediction 2)

11
Figure 3

• JNDs in orientations plotted as a function of
  obliquity.
• (A) 3 subjects
• (B) 5 subjects

12
Results

• JNDs in orientation are much smaller at
  principal orientations (horizontal or vertical)
  than at oblique orientations. (Fig. 1)
• (Replication)
• JND in orientation increases as a function of
  obliquity from the principal orientation up to
  20 obliquity and then levels off. (Fig. 3)
• (Prediction 1)
• Long stimuli give a strong oblique effect in
  orientation acuity whereas short stimuli give a
  weaker oblique effect. In other words, the
  oblique effect increases as line length
  increases. (Fig. 4)
• (Prediction 2)

13
Figure 4

• (A) JNDs in orientations plotted as a function
  of stimulus length.
• (B) Mean oblique effect index plotted as a
  function of stimulus length for 4 subjects.

14
Other explanations for these results?

• Orientations correspond to different endpoints.
  Could it be a positional instead of an
  orientational oblique effect?
• ? Replicate the experiment by just showing
  the endpoints.
• Is the effect retinal or gravitational?
• ? Replicate the experiment where subjects
  are tilted 20.

15
Results of Control Studies

• Position discrimination is much worse than
  orientation discrimination and shows no
  meridional variation. (Fig. 5)
• The meridional variation in JNDs in orientation
  follows retinal and not gravitational
  coordinates. (Fig. 6)

16
Figure 5

• JNDs in orientations for lines (open symbols) and
  equivalent differential thresholds for dots
  placed at the end of an invisible rod of the same
  length (solid symbols) at the 4 main orientations.

17
Results of Control Studies

• Position discrimination is much worse than
  orientation discrimination and shows no
  meridional variation. (Fig. 5)
• The meridional variation in JNDs in orientation
  follows retinal and not gravitational
  coordinates. (Fig. 6)

18
Figure 6

• JNDs in orientations plotted as a function of
  gravitational reference orientation for
• (A) subject R.V. in an untilted position
  (continuous line) and a tilted 20 clockwise
  position (stippled lines).
• (B) subject M.D.W. in an untilted position and a
  tilted 20 counterclockwise position.

19
Interpretation

• Both predictions came true, both control
  experiments did not threaten the conclusions. ?
• Cells similar to area 17 cells of the cat and
  monkey are involved in orientation discrimination
  for long lines.
• Of course, such a coarse correspondence between
  prediction and results does not rule out other
  alternative explanations.

20
Problems

• Logic flawed (psychophysics in people,
  physiology in animals).
• Not much contrast between the rod and the
  background, so subjects may have had a difficult
  time seeing the stimulus.
• Only 6 subjects. Some of them were the authors.
  
• Memory effects cloud clarity of conclusions.
• No control of eye movements.