Perceptual Strategy: a tale of letters and shapes

Cees van LeeuwenLaboratory for Perceptual Dynamics, KU Leuven

cees.vanleeuwen@ppw.kuleuven.beIn collaboration withThomas Lachmann

Kaiserslautern University of Technology

Overview

• Introduction: two ways in which knowledge influences perception

• Reading deficits as a consequence of learning the wrong strategy

• Differentiation in strategies between nonletters and letters

• Conclusions

Two ways in which knowledge influences perception

1. What we know influences what we perceive

The figure is determined based on familiarity. Peterson & Gibson (1993).

2. Knowledge influences how we perceive

Perceiving items as shapes vs perceiving items as letters

Habitual, as a result of learning to read!

Are Letters Special?

They are perfectly natural!

If letters are normal, why do some children have difficulty in learning to read and write?

Developmental Dyslexia

Reading Errors- Slowness of reading; problems in

understanding written materials- Losing their Line- Omissions, Repetitions, Substitution,

Insertions- Distortions of Words, Parts of Words- Reversal Errors: was ∞ saw b ∞ d

ORTON (1925) Cardinal symptom

The Causes are unknown

After more than 100 years of experimental studies

-------------------------------------------------------------------------------------------------------------Acoustic information processing VERSUS phonological deficitversus versusVisual information processing orthographical deficit

But perhaps it is both, or a more general deficit.

Thomas Lachmann’s Functional Coordination Deficit model

Not a deficit of automatization, but automatization of the wrong strategy

Our participants

57 third/fourth graders School psychologists‘ diagnosis in Grade 2Own diagnosis prior to experiment: Discrepancy Definition (2 SD Reading time SLRT vs. Raven)

The Cardinal Symptom

b vs. d vs. p vs. q

Problem with visuospatial processing of symbols?

Mental Rotation

Cooper & Shepard (1973)

Rusiak, Lachmann, Jaskowski, van Leeuwen (Perception, 2007)

N = 16 DyslexicsN = 12 Controls

R, F, G, e, k (0-180° in 45°)

“normal” oder “mirror images”?

Mental Rotation with Letters

Pressmirrored

500

700

900

1100

1300

0 45 90 135 180

Angle of Rotation

RT

(m

s)Dyslexics

Controls

Mental Rotation with Letters

- Typical Mental Rotation Effect

- Group effect (> 100 ms)

- No Interaction (no visuospatial processing deficit)

The Cardinal Symptom

b vs. d vs. p vs. q

If the problem is not with visuospatial processing, perhaps it has something to do with the structure of the configuration

SYMMETRY

Symmetry

• Makes your world simpler

• Simple patterns have few alternatives

Rotation and Reflection Symmetry Groups

Equivalence Sets (ES)Equivalence Sets Size (ESS)

Garner Patterns

Task

• Serial presentation of pairs of stimuli

ISI = 500 ms• Same or

different• Irrespective of

pattern orientation

General preference for Symmetry

Detailed model predictions possible from Degree of Pattern Symmetry (ESS) and Degree of “Sameness” (IM, CM, or NM)

Lachmann & van Leeuwen (2004) Scientific Psychology Series

Model Fit

Lachmann & van Leeuwen, 2005, Quarterly Journal of Experimental Psychology

Normals vs Dyslexics

Letters vs Shapes

Lachmann & van Leeuwen, 2007, Developm. Neuropsychology

Results

PatternsLetters

RT

(m

s)

1100

1000

900

800

700

600

symmetric

asymmetricPatternsLetters

RT

(m

s)

1100

1000

900

800

700

600

symmetric

asymmetric

Material**; Symmetry**; Group*; Material x Symmetry x Group **

Normal reading children Dyslexics

No effects on error rate No effects on error rate

Model fit*

Predicted RT (ms)

11001000900800700600

RT

(m

s)

1200

1100

1000

900

800

700

600

Dyslexics

Normal

Readers

*Note on Model fit

• Model fit was based on different assumptions for Normals and Dyslexics

• Degree of symmetry for patterns was counted for both Normals and Dyslexics

• Degree of symmetry for letters was counted for Dyslexics but ignored for Normals

• Dyslexics are faster because they give equal importance to Symmetry in Letters and Shapes

------------

Dyslexics:

• Outperform normal readers on this task• The task involves letters

• Why don’t dyslexics read better than normals?

Symmetry?

• Symmetry helps in perceiving shapes, but hinders in perceiving letters

• Strategy, based on reading practice

• When learning to read is difficult anomalous strategies

letter symmetry may be suppressed actively

Configurations

Emergent Properties (J.R. Pomerantz)

( ( (( ()

Configurations

Difficulty of recognizing the letter “E” when embedded in a Good

figure

Configurations

Some Congruence Effects

GREEN

RED

Push the button on the

Right

Letters

versus

Shapes

Letters vs Shapes

Congruent versus Incongruent

Van Leeuwen & Lachmann, Perception & Psychophysics, 2004

435

440

445

450

455

460

465

470

475

480

Letters Shapes

CongruentIncongruent

Binary Classification

Van Leeuwen & Lachmann, Perception & Psychophysics, 2004

410

420

430

440

450

460

470

480

490

Letters Rotated Letters Pseudo-Letters

CongruentIncongruent

Binary Classification

Binary Classification

• Rotated letters Pseudo-letters, and Geometrical Shapes show Congruence Effects.

• Letters show Negative Congruence Effects.

• Is the effect related to the response alternatives?

Contrasting Response Categories

Similarity

Press Left Button: C or Circle or L or Rectangle

Press Right Button: H or Square or A or Triangle

No Similarity

Press Left Button: C or Square or L or Triangle

Press Right Button: H or Circle or A or Rectangle

Similarity

0

100

200

300

400

500

600

Letters Shapes

CongruentIncongruent

455

460

465

470

475

480

485

490

495

Letters Shapes

CongruentIncongruent

Contrasting Response Categories

No Similarity

• What about less experienced readers?• What about developmental dyslexics?

In normal adult readers:

Dissociation in Early Visual IntegrationProcesses of Letters and Shapes

Binary Classification

Normal ReadingChildren

537541 553531 546547N =

MATERIAL

shapeletter

95

% C

I R

T

560

540

520

500

480

460

CONINCON

iso

con

incon

648642 651620 669638N =

MATERIAL

shapeletter

95

% C

I R

T

600

590

580

570

560

550

540

530

520

CONINCON

iso

con

incon

• Weaker dissociation in normal children than in adults

• Dyslexics have increased response times and variance AND an adult dissociation pattern

PUZZLE

DyslexicChildren

RT (ms)

RT (ms)

Puzzle

Normal reading children have a weaker dissociation than adults strategy differentiation not yet well established?

Dyslexics probably do not have more reading experience dissociation should be same or even weaker

Quod Non!

Puzzle

Our Dyslexics sample has:

-slow response rates

-strong dissociation

Does this mean that all individuals in the sample have both these characteristics?

Puzzle

Perhaps:

-slow response rates Subpopulation I

-strong dissociation Subpopulation II

Subtypes of Dyslexia?

Analogy with Lachmann (2003, 2005)

Puzzle

Diagnostic subgroups

Reading test: SLRT (Landerl, Wimmer, & Moser, 1997) word and non-word reading subtests

Non-Word Reading Impaired (NWRI):at least reading time in non-word reading > 2 SD

Frequent Word Reading Impaired (FWRI):reading time in frequent word reading > 2 SD, but normal non-word reading time (within 1 SD)

Puzzle

FWRI NWRI

Fuchs &Lachmann (2003)

Lachmann (2007)

Lachmann et al.(2005)

Delayed auditory/visualsyllable discrimination

Delayed letter identification

Attenuated mismatchnegativity syllables

Pronounced fast-same effect in visual conditions

Normal mismatch negativity

Large modality effects

increased response times but normal between conditions effects

specific effects enhancedmostly involvingvisuo-auditory coordination

Same-Different Task• Sequential presentation of

two items

• First always isolated, second isolated or in congruent or incongruent surroundings

• Letters, Pseudo-letters, Shapes; No mixed-category pairs

• Adults, normal reading children, FWRI vs NWRI dyslexics

Material x congruence x group.

Group MaterialCongruence

Isolated Congruent Incongruent

Normal reading adults

Letters 478 483 483

Pseudo-letters 501 498 515

Shapes 459 482 519

Normal reading Children

Letters 690 673 722

Pseudo-letters 698 737 771

Shapes 694 724 746

FWRI-dyslexics

Letters 1006 965 1059

Pseudo-letters 990 1015 1085

Shapes 1016 1035 1122

NWRI-dyslexics Letters 745 836 755

Pseudo-letters 792 797 774

Shapes 712 784 842

Summary

• No differences in visuospatial manipulation• Dyslexics sometimes even better than Normals

• Letter/Non-letter Dissociation

• Anomalous in Dyslexics

• Subgroups can be identified:

-- “General Slowing down” in FWRI

-- Enhanced Dissociation in NWRI

Hot from the PressFernandes, T., Vale, A.P., Martins, B., Morais, J., & Kolinsky, R. (2014) Processing in

Developmental Dyslexia: Combining evidence from dyslexics, typical readers and illiterate adults. Developmental Science.

To clarify the link between anomalous letter processing and developmental dyslexia, we examined the impact of surrounding contours on letter vs. pseudo-letter processing by three groups of children – phonological dyslexics and two controls, one matched for chronological age, the other for reading level – and three groups of adults differing by schooling and literacy – unschooled illiterates and ex-illiterates, and schooled literates. For pseudo-letters, all groups showed congruence effects (CE: better performance for targets surrounded by a congruent than by an incongruent shape). In contrast, for letters, only dyslexics exhibited a CE, strongly related to their phonological recoding abilities even after partialling out working memory, whereas the reverse held true for the pseudo-letter CE. In illiterate adults, the higher letter knowledge, the smaller their letter CE; their letter processing was immune (to some extent) to inference from surrounding information. The absence of a letter CE in illiterates and the positive CE in dyslexics have their origin in different aspects of the same ability, i.e. phonological recoding.

A straightforward replication but with more straightforward results

Conclusion

Possible mechanisms:

•Letter treated same as non-letter shapes?

•Problems with surrounding noise suppression?

Thanks to:

• Prof. Thomas Lachmann, TU Kaiserslautern

and his students: Jessika Wüst, Katrin Berg, Sabine Jenke, Sandra Schlademann, Patricia Rusiak, Dave Barton e.a.

• Third Primary School Leipzig

…and to you for your patience!