An Electrophysiological Study of Object-Based Correspondence Effects:
Is Visual Processing Modulated by the Intended Action?
Elliott Jardin & Mei-Ching Lien
Robert W. Proctor
Oregon State University
Purdue University
Results and Discussion
0.12
-2 ms
Proportion of Error
2 ms
600
575
550
525
500
.006
0.10
0.08
0.06
0.04
0.02
0.00
Kitchen
Tool
Kitchen
Category
-2
0
100
200
300
400
500
-200
0
100
200
300
400
500
0
N1
2
uv
uv
0
-100
N1
2
600
-1 ms
550
525
500
Category
4
.005
6
6
0.04
0.02
EEG Data:
P1
Fixation
1000-1200 ms
Target
2000 ms
Feedback Tone
100 ms
Time
ERP Measures: EEG epochs were time-locked to stimulus onset,
relative to a 200-ms pre-stimulus baseline. The P1 effect was measured
100-130 ms after stimulus onset and the N1 effect 150-200 ms after
stimulus onset, averaged across the P7, P8, O1, O2 electrode sites.
Predictions
26 ms**
Kitchen
Tool
575
550
525
500
Category
0.10
0.08
.022**
.023**
Kitchen
Tool
0.06
0.04
0.02
Tool
Category
Central Location
Peripheral Location
Post-Stimulus Interval (ms)
-100
N1
0
100
The relative location coding view predicts the correspondence effect
should be absent for central object but present for peripheral objects,
whereas the grasping affordance view predicts the presence of the
correspondence effect regardless of the object location.
The grasping affordance account predicts correspondence effects in
behavioral data (e.g., RT) for both kitchen utensils and tools, as well as
in the ERP data (P1 and N1 amplitude).
200
300
400
500
-200
-100
0
0
0
2
2
4
4
P1
6
100
200
N1 300
400
500
P1
6
Post-Stimulus Interval (ms)
Post-Stimulus Interval (ms)
Experiment 2 (N=20)
Averaged across kitchen
utensils and tools
-2
-200
-100
-2
N1
0
100
200
300
400
500
-200
-100
0
0
0
2
2
uv
Exp2 examined whether the absence of correspondence effects in Exp1
was due to the lack of location coding for the grasping component. The
object (against a white background) was presented centrally, as in Exp1,
or peripherally (5.19° from central fixation) so the handle was clearly
located to the left or right. We used only 8 kitchen utensils and 8 tools
with distinct bases and handles. Object location (central vs. peripheral)
was randomly intermixed within blocks. For the peripheral objects, the
handle orientation (left vs. right) and object location (left vs. right)
always corresponded.
-2
uv
uv
The correspondence effects on RT were negligible for kitchen utensils
and tools, Fs<1.0. Also, the congruency between handle orientation
and response location had no effect on P1 and N1 for either category,
Fs<1.0. These findings are inconsistent with Goslin et al.’s (2012) claim
and suggest that visual attention is not modulated by object affordance.
-200
uv
+
24 ms**
0.00
0.00
-2
Event Sequence:
600
Category
Post-Stimulus Interval (ms)
Stimuli: Goslin et al.’s (2012) 84 pictures – 42 kitchen utensils and 42
tools – were each repeated 6 times. Within a category, half the handles
were on the left and half on the right. Handle orientation (left vs. right)
was congruent or incongruent with response location (left vs. right).
-.002
0.06
4
P1
625
0.12
0.10
0.08
Peripheral Location
650
Tool
Kitchen
Experiment 1 (N=21)
Task: Participants made a speeded keypress indicating whether a
centrally located object was a kitchen utensil or a tool. The mapping
varied between blocks for each participant.
7 ms
575
0.12
Proportion of Error
-100
625
Kitchen
Averaged across kitchen
utensils and tools
-2
*p<.05, **p<.001
Central Location
650
Tool
Category
EEG Data:
-200
.008*
Response Time (in ms)
*p<.05, **p<.001
650
625
Behavioral Data:
Response Time (in ms)
Behavioral Data:
Response Time (in ms)
People often respond faster when the orientation of an object’s
graspable part corresponds with the response location than when it
does not – the object-based correspondence effect (e.g., Tucker & Ellis,
1998). Some attribute the effect to relative location coding (e.g., Cho &
Proctor, 2011), whereas others attribute it to grasping affordance (e.g.,
Goslin et al., 2012). In Goslin et al.’s study, for instance, participants
categorized a centrally located object (kitchen utensil vs. tool). The
handle orientation (left vs. right) did or did not correspond with the
response location (left vs. right). Using event-related potential (ERP)
measures, they found that early visual processing, reflecting on the P1
and N1 ERP components, was modulated by the grasping action
afforded by the object. One concern regarding their findings, however,
is that the correspondence effect on behavioral data (response time
[RT]) was negligible (only 10 ms for tools and absent for kitchen
utensils). The present study therefore tested Goslin et al.’s affordance
account against the location coding account of correspondence effects.
Results and Discussion
Proportion of Error
Introduction
4
6
4
P1
100
N1 300
200
400
500
P1
6
Post-Stimulus Interval (ms)
Post-Stimulus Interval (ms)
As in Exp1, central objects showed little correspondence effects on RT
and PE, Fs(1,19)≤3.10, ps≥.09. However, strong effects were observed
for peripheral objects, Fs(1,19)≥17.28, ps≤.001. The P1 and N1 were
modulated by the congruency between handle orientation and
response location for peripheral objects but not central objects,
Fs(1,19)≥4.88, ps≤ 0.05. These suggest that correspondence effects are
due to the location coding of objects, not the grasping orientation.
General Discussion
Central Location
Peripheral Location
References
Cho, D., & Proctor, R. W. (2011). Correspondence effects for objects with opposing left and right protrusions. Journal of Experimental Psychology: Human Perception and Performance, 37, 737-749.
Goslin, J., Dizon, J., Fischer, M. H., Cangelosi, A., & Ellis, R. (2012). Electrophysiological examination of embodiment in vision and action. Psychological Science, 23, 152-157.
Tucker, M., & Ellis, R. (1998). On the relations between seen objects and components of potential actions. Journal of Experimental Psychology: Human Perception and Performance, 24, 830-846.
We tested the claim that correspondence effects are the result of object
affordance. We found correspondence effects on behavioral data (RT
and PE) and ERP data (P1 and N1) only when objects were peripherally
located. These findings contradict Goslin et al.’s (2012) findings and
their hypothesis that intended grasping action modulates visual/spatial
attention. Instead, they support the relative location coding account.