Introduction Mario Fifić , Daniel Little

advertisement
A Snake Wiggle of Reaction Time
Functions to Indicate Holistic Perception
Old face
Configurally Features only
Altered
(new) face
Mario Fifić1, Daniel Little2
1Grand Valley State University, USA
2The University of Melbourne, Australia
Phenomenon
Introduction
Errors
RT(ms)
Participant 01
780
MeanParticipant
RT trend
02
900
760
700
0.2
680
660
640
0.1
620
560
0
2
4
6
8
10
400
12
0
Participant 03
600
0.1
HL
0
2
4
6
8
HL
10
Lips- position
H
0
LL2 L
0.0
4
8
10
12
Serial model RT[Ext]<RT[Int]
Jets Parallel model RT[Ext]=RT[Int]
Participant Coactive
06
RT[Ext]>RT[Int]
Block (2 x Session)
Eye- separation
0.3
540
760
20
0
RT[Ext]-RT[Int]
0.1
0.0
1
2
3
4
5
6
7
0.2
720
-20
700
-40
680
-60
660
-80
640
8
“Coactive”
740
1
2
3
4
5
6
7
8
9
10
0.1
0
-100
Proportion of errors
Interior
RT (ms)
0.2
560
Proportion of errors
RT (ms)
6
(Fific, Little & Nosofsky, 2010; Psych Review)
500
2
4
6
8
Snake Wiggle
0.0
“Parallel”
10
Block (2 x Session)
Conclusions
Serial
Self-terminating
Input
Eyes
Lips
Decision
OR
Response
B
Serial
Exhaustive
Input
Eyes
Lips
Decision
AND
Response
Input
Parallel
Self-terminating
Input
D
Parallel
Exhaustive
Eyes
Decision
OR
Response
Lips
Input
Eyes
Input
Lips
Decision
AND
Response
E
(Fific & Townsend, 2010)
Interior
780
0.3
520
Architecture flow
diagram
Eye- separation
0.1
580
Exterior
580
A catalog of mental architectures
Contrast Category
LH
640
800
600
A
Jets
660
Errors
RT(ms)
620
Analytic
Analytic/
Holistic
C
H
Exterior
Sharks
LL
620
640
Block (2 x Session)
LL L
HH
HL
680
Participant 05
Sharks
LL
HL
Contrast
0.2
Category
700
LH
600
Sharks 0.0
550
A Catalog
Sharks
0.3
Proportion of errors
650
HH
RT (ms)
700
Proportion of errors
0.2
Lips- position
RT (ms)
HH
SIC(t) = Sll(t) - Slh(t)- (Shl(t) - Shh(t))
HL
Participant 04
720
Target Category
LH
0
LH
10
740 classification “AND”
Conjunctive-rule
0.3
Survivor interaction contrast (SIC):
HH
HL
8
760
800
The main SFT statistic
6
Converging evidence
850
Tutorial
4
Errors
RT(ms)
460
LH
2
Block (2 x Session)
480
HH
0.0
Block (2 x Session)
Parallel, or coactive architecture
Mandatory exhaustive stopping rule
Super-capacity
Interdependencies between feature
detectors
Target
LH
HH Category
600
500
0.0
Block (2 x Session)
Conjunctive-rule classification “AND”
0.2
580
Defining holism/configurality in
terms of processing characteristics:
Systems factorial technology (SFT)
Experimental study
700
0.1
600
750
•
•
•
•
800
RT (ms)
RT (ms)
720
0.3
Proportion of errors
0.3
740
Proportion of errors
We analyzed the underlying fundamental processes engaged in
forming holistic perceptual representations. The subjects
participated in a face categorization task over multiple
sessions. We applied the systems factorial technology (SFT) to
analyze the properties of the observed response time (RT)
distributions. The key statistic was a survivor interaction
contrast function (SIC). Over the course of extensive practice,
the observed SICs exhibited a specific pattern of shape
transformations that could be described as a "snake wiggle".
The observed SIC signature indicated that the processing
mechanism behind holistic perception relies on strong positive
facilitation between feature detectors, within the parallel
mental network. The converging evidence is provided by the
additional qualitative RT test (Fific, Little & Nosofsky, 2010).
Coactive
Input
Input
Eyes
Lips
Joe’s
face
Decision
Response
Analytic
Analytic/
Holistic
Strong
Holistic
Signatures
 Let’s wiggle: The SIC function wiggles
its way to S-shaped positive function
 Robustness: All subjects wiggled
 Coactivation: The target signature is
Coactive, an indicator of strong holism
 Performance superiority: Mean RT
decreases, and accuracy improves
 Introspection: The verbal reports
 Contrast faces wiggled to coactivation,
too.
 The snake wiggle is a dynamic signature
of holistic perception
Recommended Readings
Fific, M., Little, D. R., & Nosofsky, R. M. (2010). Logical-rule models of classification response
times: A synthesis of mental-architecture, random-walk, and decision-bound approaches.
Psychological Review, 117, 309–348.
Fific, M., & Townsend, J. T. (2010). Information-processing alternatives to holistic perception:
Identifying the mechanisms of secondary-level holism within a categorization paradigm.
Journal of Experimental Psychology: Learning, Memory, and Cognition, 36, 1290-1313.
E-mail: fificm@gvsu.edu
Download