Does drawing faces make you a super-expert of faces? An investigation of face perception and recognition abilities in visual artists.

(1)

M

Methods

Does drawing faces make you a super

Does drawing faces make you a super----expert of faces?

expert of faces?

An investigation of face perception and recognition abilities in visual artists

B

Background

Face processing abilities would range from very poor (i.e. developmental prosopagnosia) to outstanding performance (i.e. super-recognizers)1_{. Highly physionomist}

people seem to show a larger face inversion effect compared to controls, suggesting a link between face recognition skills and the amplitude of the effect. Through the practice of portraiture and the close attention paid to the shape of the head, the facial features and their configuration to reach a good likeness of a

person, artists might perform like « super-recognizers »

.

A

Aim

Investigate whether the regular practice of portraiture leads to specific better performance in face processing (perception and memory) and whether the expertise acquired through drawing qualitatively affects the processing of faces (i.e. as indexed by a differential inversion effect).

P PP

Participants

N = 21; 11 artists (M= 26 years ± 4), 10 controls (M= 26 years ± 3)

Artists drew portraits for 8,5 years ± 5

Mental rotationMental rotationMental rotationMental rotation

Same – different judgment on pseudo 3D figures4_{presented side by side}

Rotation on the X (50 trials) or on the Z axis (50 trials)

Perceptual matching of Houses vs. Faces Perceptual matching of Houses vs. Faces Perceptual matching of Houses vs. Faces Perceptual matching of Houses vs. Faces (Upright/inverted) Image 1 (1000 ms) → Mask (750 ms) → Image 2 (till response)

Houses Houses Houses

Houses FacesFacesFacesFaces

“Same” pair (36 trials) “Same” pair (36 trials)

Artists drew portraits for 8,5 years ± 5

They devoted 62% ± 23 of their practice to faces and 33% ± 18 to portraits/caricatures

M MM

Measures and tasks

SelfSelfSelfSelf----reported face processing abilities questionnaire reported face processing abilities questionnaire reported face processing abilities questionnaire reported face processing abilities questionnaire 37 items, 7-point Likert scale (e. g. “I recognize people I have only encountered once at first sight”)

Cambridge Face Perception Cambridge Face Perception Cambridge Face Perception Cambridge Face Perception TestTestTestTest2222

Arrange 6 faces according to their degrees of likeness to a target

Cambridge Face Memory Test Cambridge Face Memory Test Cambridge Face Memory Test Cambridge Face Memory Test ––––Australian³Australian³Australian³Australian³

Retrieve 1 out of 6 studied targets among 3 more and more degraded faces

R

Results & D

D

Discussion

D

Face

Face Face

Face processingprocessingprocessingprocessing questionnaire: questionnaire: questionnaire: questionnaire: Artists reported better face recognition abilities (M = 204) than controls (M = 159); Mann-Whitney U test, z = 3, p < 0.005.

0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

Intro No noise Noise Total

CFMT CFMT CFMT

CFMT –––– AustralianAustralianAustralianAustralian ((((accuracyaccuracyaccuracyaccuracy))))

Artists - Up Artists - Inv Controls - Up Controls - Inv

the Z axis (50 trials)

5 angles (0, 45, 90, 135 and 180°)

“Same” pair

“Different” pair

Perceptual matching task Perceptual matching task Perceptual matching task Perceptual matching task

0 0,2 0,4 0,6 0,8 1

Faces Houses Faces Houses Artists Controls A c c u ra c y A c c u ra c y A c c u ra c y A c c u ra c y Upright Inverted

Mental rotation task Mental rotation task Mental rotation task Mental rotation task

0,6 0,7 0,8 0,9 1 0 45 90 135 180 A c c u ra c y A c c u ra c y A c c u ra c y A c c u ra c y Rotation angle Rotation angle Rotation angle Rotation angle in degreesin degreesin degreesin degrees

Artists X Artists Z Controls X Controls Z

“Different” pair (36 trials) “Different” pair (36 trials)

R

References

1. Russell et al. (2009). Super-recognizers: People with extraordinary face recognition ability.

Psychonomic Bulletin & Review, 16, 252-257.

2. Duchaine et al. (2007). Family resemblance: Ten family members with prosopagnosia and within-class object agnosia. Cognitive Neuropsychology, 24, 419-430.

3. McKone et al. (2011). Face ethnicity and measurement reliability affect face recognition performance in developmental prosopagnosia: Evidence from the Cambridge Face Memory Test–Australian.

Cognitive Neuropsychology, 28, 109-146.

4. Peters & Battista (2008). Applications of mental rotation figures of the Shepard and Metzler type and description of a Mental Rotation Stimulus Library. Brain and Cognition, 66, 260-264.

5. Kozbelt (2001). Artists as experts in visual cognition . Visual Cognition,8, 705-23.

C

Contact information: cdevue@ulg.ac.be

C

Conclusion

Artists who draw portraits on a regular basis report better face recognition abilities. They also show better face perception skills as well as better perceptual skills in general (see 5) in terms of accuracy, which sometimes relied on longer processing times (see perceptual matching task). These better perceptual abilities extend to better face recognition abilities. The absence of differential inversion effect in artists might indicate a quantitatively but not

qualitatively different facial processing due to expertise in this group compared to controls.

Intro No noise Noise Total

2 Group X 2 Difficulty (No noise, Noise) X 2 Orientation Main effects of Group (artists > controlsartists > controlsartists > controlsartists > controls), Difficulty and

Orientation, all ps < 0.05 (no interactions)

0 10 20 30 40 50 60 70 80 Upright Inverted CFPT CFPT CFPT

CFPT ---- Mean total deviationMean total deviationMean total deviationMean total deviation

Artists

Controls

2 Group X 2 Orientation

Main effects of Group (artists > controlsartists > controlsartists > controlsartists > controls) and Orientation, both ps < 0.01 (no interaction) Similar inversion effect in the two groups ((upright

-inverted)/upright): -1.1 ± 0.75 vs -0.95 ± 0.63 Artists have better perceptual and recognition abilities Artists have better perceptual and recognition abilities Artists have better perceptual and recognition abilities Artists have better perceptual and recognition abilities

for faces than controls for faces than controls for faces than controls for faces than controls

2 Group X 2 Category X 2 Orientation

Main effects of Group (artists > controlsartists > controlsartists > controlsartists > controls), Category (houses > faces) and Orientation (upright > inverted), all ps < 0.001

Category X Orientation interaction (inversion effect only inversion effect only inversion effect only inversion effect only for faces

for faces for faces for faces)

Main effects of Group (artists > controlsartists > controlsartists > controlsartists > controls) and Orientation (upright < inverted), all ps < 0.05 Category by Orientation interaction (inversion effect inversion effect inversion effect inversion effect only for faces

only for faces only for faces

only for faces), Group X Category interaction (controls faster with houses than with faces)

Artists are overall more accurate but slower than controls; Artists are overall more accurate but slower than controls; Artists are overall more accurate but slower than controls; Artists are overall more accurate but slower than controls; they do not show an ampler face inversion effect

they do not show an ampler face inversion effect they do not show an ampler face inversion effect they do not show an ampler face inversion effect

0 500 1000 1500

Faces Houses Faces Houses Artists Controls M e a n R T s ( m s ) M e a n R T s ( m s ) M e a n R T s ( m s ) M e a n R T s ( m s ) Upright Inverted

2 Group X 2 Axis X 5 Rotation angle Main effects of Group (artists > controlsartists > controlsartists > controlsartists > controls), and angle, all ps < 0.05

Axis X angle interaction, p< 0.002

Main effects of axis (X > Z), and angle, all ps < 0.001

Axis by angle interaction, p< 0.002

Similar processing time for artists and controls Similar processing time for artists and controls Similar processing time for artists and controls Similar processing time for artists and controls but artists are more accurate

but artists are more accurate but artists are more accurate but artists are more accurate

Rotation angle Rotation angle Rotation angle Rotation angle in degreesin degreesin degreesin degrees

2000 3000 4000 5000 6000 7000 8000 9000 0 45 90 135 180 M e a n R T s M e a n R T s M e a n R T s M e a n R T s ( m s ) (m s ) (m s ) (m s )

Rotation angle in degrees Rotation angle in degrees Rotation angle in degrees Rotation angle in degrees

Artists X Artists Z Controls X Controls Z