how does attention affect perception?
DESCRIPTION
How Does Attention Affect Perception?. Dr. Linda Rueckert presented to the Northeastern Illinois University Psychology Dept., November, 2006. Which side of the line is shorter?. The Human Visual System I. The Human Visual System II. The Right Cerebral Hemisphere. Left Hemi-neglect. - PowerPoint PPT PresentationTRANSCRIPT
How Does Attention Affect Perception?
Dr. Linda Rueckertpresented to the Northeastern Illinois University Psychology
Dept., November, 2006
Which side of the line is shorter?
The Human Visual System I
The Human Visual System II
The Right Cerebral Hemisphere
Left Hemi-neglect
Left Hemi-neglect Most common after right parietal lesions. Line bisection task:
Patients with neglect transect the line to the right. (i.e. they underestimate the left side of the line.)
Error is greater for longer lines Is this a sensory or motor phenomenon?
“Landmark” Task
“Which side of the line is smaller (left or right)?”Or
“Which side of the long line is the short line closer to?”
Neglect patients say the left side is shorter.(i.e. they underestimate the left side)
Results from Previous Studies Using Short Lines (“cross-over” effect)
Neglect patients put the cross-mark too far left.(Overestimate left side)
Neglect patients say the right side is shorter(Overestimate left side)
Line Bisection:
Landmark:
Normal Subjects and Long Lines Line bisection:
Normal subjects transect slightly to the left. (i.e. they underestimate the right side)
Landmark task:
Normal subjects say the right side is shorter. (i.e. they underestimate the right side)
Will normal subjects show a cross-over effect with short lines? Rueckert, Deravanesian, Baboorian,
Lacalamita & Repplinger, 2002.
Subjects tested on both the line bisection and Landmark tasks.
Lines were 1, 2, 8, 16 or 28 cm long.
Results from Normal Subjects on Landmark Task
30201000.3
0.4
0.5
0.6
0.7
0.8
Line Length (cm)
Pro
port
ion L
eft
Choic
es
Results Normal subjects did show the cross-over
effect on the Landmark task (i.e. they overestimated the right side of short lines).
They did not cross-over on the line bisection, due to perfect performance on short lines.
Will normal subjects show a cross-over effect on line bisection if the task is made more difficult?
Results from Normal Subjects on Timed Line Bisection Task
20151050-2.0
-1.6
-1.2
-0.8
-0.4
0.0
Line Length (cm)
Err
or
(mm
)
Conclusions Normal subjects show a cross-over effect
on the Landmark task, but not on regular line bisection.
The cross-over appears to be a perceptual phenomenon, as opposed to a motor phenomenon.
Unanswered Questions Is the cross-over point determined by the
absolute (actual) or relative length of the line?
If it’s relative, relative to what?
Context Effects in Neglect Patients Marshal et al. (1998) – found the cross-
over point was determined by the relative length of all lines in a block of testing.
Ricci & Chatterjee (2001) – found cross-over point was not entirely determined by context.
Context Effects in Normal Subjects
Rueckert and McFadden, 2004
Administered Landmark task using 1 to 20 cm lines.
20 participants tested on only 1, 2, and 8 cm lines.
20 participants tested on only 8, 16, and 28 cm lines.
Context Effects in Normal Subjects: Results
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1 2 8 16 28
Line Length (cm)
Pro
portio
n le
ft ch
oice
sxxx
x
Short group
Long group
Context Effects: Conclusions
Normal subjects are less affected by context than neglect patients are.
Summary of Landmark Results inPatients and Controls
Neglect Patients Normal subjects
Long Lines say too far L say too far R
(underestimate L side) (underestimate R side)
Short Lines say too far R say too far L
(overestimate L side) (overestimate R side)
The Model1. There is a normal tendency to
overestimate the length of short lines and underestimate the length of long lines (the regression effect; Hollingworth, 1909 , Huttenlocher et al., 2000; Werth, & Poppel, 1988).
2. This tendency is greater when attention levels are low.
Neglect patients should have a greater tendency to underestimate long lines and overestimate short, due to lower levels of attention (Mennmeier, et al., 2002; Tegner, & Levander, 1991).
This effect should be greater in their left visual field, resulting in underestimating the left side of long lines, and overestimating the left side of short lines, resulting in the cross-over effect.
This also explains the cross-over effect found with weights (Chatterjee et al., 2000)
Since normal subjects pay less attention to the right side of space, they would underestimate the length of the right side of long lines, and overestimate the length of the right side of short lines.
The Model1. There is a normal tendency to
overestimate the length of short lines and underestimate the length of long lines (the regression effect; Hollingworth, 1909 , Huttenlocher et al., 2000; Werth, & Poppel, 1988).
2. This tendency is greater when attention levels are low.
Does lack of attention lead to anunderestimation of long lines and overestimation of short lines in normalsubjects?
Rueckert, 2003: Method Lines 1 – 20 cm long flashed on a
computer screen.
Subject has to draw a line of the same length.
During some blocks of trials attention is reduced by requiring subjects to remember a set of 6 numbers.
Results Subjects overestimated the length of short
lines and underestimated the length of long lines.
This effect was greater in the attentional load condition.
Rueckert 2006: Method 1, 1.5, 2, 4, 6, 8, 12, 16 and 20 cm
vertically-oriented lines were presented in the LVF or RVF.
On some trials an arrow preceded the line.
For most trials, the arrow correctly told the subject where the line would appear (validly cued trials).
For a few trials the arrow appeared in the opposited field (invalidly cued trials).
Predicted Results Subjects would overestimate the length of
short lines and underestimate the length of long lines.
This effect would be greater for lines in the right field.
This effect would be greater for invalidly cued trials.
Main Effect of Line Length
-0.2
0
0.2
0.4
0.6
0.8
Short Long
Line Length
Pro
pn. E
rror
(Ove
rest
imat
ion)
Validity by Visual Field: Short Lines Only
0.4
0.45
0.5
0.55
0.6
0.65
0.7
0.75
0.8
LVF (R hemisphere) RVF (L hemisphere)
Valid
Invalid
Pro
pn.
Err
or (
over
estim
atio
n)
Conclusions The predicted results were found for short
lines only. Short lines were overestimated, especially
when attention was low (i.e. in the RVF and on invalidly cued trials).
Overall Conclusions The cross-over effect is well established in both
normal subjects and neglect patients.
The proposed model can explain the effect in both groups.
Additional evidence is needed to determine whether the regression is enhanced when attention levels are low.
Further study is needed to determine the effect that perception of length is affected by context.