print-to-sound not print-to-meaning training helps ... · 24 adults each learn to read two...

1
Plaut, D. C., McClelland, J. L., Seidenberg, M. S., & Patterson, K. (1996). Understanding normal and impaired word reading: Computational principles in quasi-regular domains. Psychological Review, 103(1), 56-115. Taylor, J. S. H., Plunkett, K., & Nation, K. (2011). The influence of consistency, frequency, and semantics on learning to read: An artificial orthography paradigm. Journal of Experimental Psychology: Learning Memory and Cognition, 37(1), 60-76. Taylor, J. S. H., Rastle, K., & Davis, M. H. (2013). Can cognitive models explain brain activation during word and pseudoword reading? A meta-analysis of 36neuroimaging studies. Psychological Bulletin, 139(4), 766-791. Taylor, J. S. H., Rastle, K., & Davis, M. H. (2014). Distinct Neural Specializations for Learning to Read Words and Name Objects. Journal of Cognitive Neuroscience, 26(9), 2128-2154. Print-to-sound not print-to-meaning training helps decoding and comprehension: Orthographic learning and fMRI J. S. H. Taylor 1 , Matthew H. Davis 2 , and Kathleen Rastle 1 1 Royal Holloway, University of London, UK 2 MRC Cognition and Brain Sciences Unit, Cambridge, UK 1. Background 2. Method 5. Asymmetric differences in brain activity 6. Conclusions 7. References 3. Learning to read aloud and comprehend 4. Print–sound training benefits comprehension Grant Number: ES/L002264/1 df rs h ij uv fɒⅿ” See Say See Say “camel” df rs h ij uv Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9 Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9 Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9 Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1 2 3 4 5 7 8 6 Number of mes aempted Number of mes aempted 1 2 3 4 5 7 8 6 Day 1 Day 2 Day 3 Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9 Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9 Day 1 Day 2 Day 3 Orthography-to-Phonology Focus Orthography-to-Semantic Focus 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 1000 2000 3000 4000 5000 Reading Aloud: Proportion Items Correct (+/- 1SE) Response Time in ms (+/- 1SE) Transfer Questions (Box 4) Does print-sound training improve comprehension? Does print-meaning traning improve reading aloud? phonology semantics orthography print-to-sound-to-meaning pathway print-to-meaning pathway Models of reading propose two ways to comprehend written text Plaut et al., 1996 Corroborated by brain imaging data Taylor, Rastle, & Davis, 2013 L intraparietal cortex L posterior occipitotemporal cortex L inferior frontal gyrus L anterior fusiform/ temporal cortex Simple Questions (Box 3) Does print-sound training improve reading aloud? Does print-meaning traning improve comprehension? fаɪⅿ fɒⅿ bᴂv ꜱⅰ:f Orthography-to-Phonology (O-P) focus language Orthography-to-Semantics (O-S) focus language bc ed qr z a ij uv st fg st z jk vwx df rs h ij uv Print Sound Meaning 24 adults each learn to read two different artificial orthographies, each consisting of 24 novel words Manipulate focus of training within subjects: Orthography-to-Phonology focus: for one orthography, more training on systematic print-sound mappings Orthography-to-Semantics focus: for other orthography, more training on arbitrary print-meaning mappings Reading aloud: Spelling: Rhyme Judgement: Comprehension: Orthographic Choice: Semantic Categorisation: 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 1000 2000 3000 4000 5000 Comprehension: Trained Untrained Trained Untrained Generalisation: Day 5 Day 10 Orthography-to-Phonology focused training benefits accuracy and speed of reading aloud. Orthography-to-Phonology Focus Orthography-to-Semantic Focus If we equate practice on the specific task, does learning transfer to the alternative mapping? Orthography-to-Semantic training does not transfer: Days 1-4, 6-9 also involved 3 x print-meaning tasks. This has no benefits for reading aloud. Instead, reading aloud accuracy depends only on practice with print-sound mappings. Within regions active for [pseudowords - words] in English Implicated in print-to-sound mapping [Orthography-Semantic > Orthography-Phonology Focus] Differences in activity during reading aloud p < .001 uncorrected, p < .05 FWE cluster corrected No difference in activity when saying word meanings x = -50 x = -40 x = -20 phonology semantics orthography Participants received pre-training in phonology-semantic mappings (establish an oral vocabulary). Following this, we found that: Orthography-Phonology training was beneficial for both reading aloud and comprehension. Orthography-Semantic training was a slower way to learn to comprehend and was detrimental for reading aloud as well as generalising to untrained words. Thus, even with minimal oral vocabulary training, it was easier for learners to use print-sound knowledge to access word meanings, than to map from print-to-meaning directly. Our precisely controlled experiment therefore supports existing UK teaching practice, which focuses on phonics (sounds and letters) in the early years. Asymmetric benefits of print-sound and print-meaning training 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 x = -50 x = -20 x = -40 No significant differences in activity Reading aloud is more effortful following orthography-semantic than orthography-phonology training posterior occipitotemporal cortex precentral gyrus intraparietal sulcus angular gyrus anterior middle temporal gyrus superior frontal gyrus No difference in effort of saying meanings following the two types of training fɒⅿ” 0.7 0.75 0.8 0.85 0.9 0.95 1 1000 1500 2000 2500 3000 3500 Day 5 Day 10 See trained word Say pronunciation df rs h ij uv Say meaning “camel” Proportion Items Correct (+/- 1SE) Response Time in ms (+/- 1SE) Proportion Items Correct (+/- 1SE) Response Time in ms (+/- 1SE) See untrained word Say pronunciation “ⅿɒvSee trained word Trained Untrained Trained Untrained Orthography-to-Semantic focused training benefits speed but not accuracy of saying meanings. Orthography-to-Phonology focused training benefits accuracy and speed of generalisation to untrained words. Proportion Items Correct (+/- 1SE) Proportion Items Correct (+/- 1SE) Orthography-to-Phonology training transfers: Days 1-4, 6-9 also involved 3 x print-sound tasks. This transfers and improves saying the meaning. So comprehension accuracy depends more on print-sound than print-meaning practice. Comprehension: df rs h ij uv Say meaning “camel” See trained word Reading Aloud: fɒⅿ” See trained word Say pronunciation Day 0 Sound-Meaning pre-exposure Days 1-4 and 6-9 Print-to-sound tasks O-P focus language x3 per day O-S focus language x1 per day Print-to-meaning tasks O-S focus language x3 per day O-P focus language x1 per day Day 11: fMRI scan Reading aloud Saying meanings English word reading See Say See Say “camel” Hear Select - - Hear Select df rs h ij uv See Select df rs h ij uv from all 24 a ij uv st df rs h ij uv kl xy bc bc See Select odd-one-out vehicle animal animal Say “axe” Hear See Say Read aloud English words and pseudowords Sparse imaging design - silent intervals between scans [Orthography-Semantic vs. Orthography-Phonology Focus] Training procedure Task Details x3 per day x1 per day pre-train Print Sound Meaning x1 per day x3 per day pre-train O-P: O-S: See Say O-S: O-P: See Say “tractor” Within regions active for [words - pseudowords] in English Implicated in print-to-meaning mapping Days 5 and 10 Test Sessions ḃəʊꜱ �еꜱ gʌ� vū:z gʌ�” fаɪⅿ” fɒⅿ” fɒⅿ” “lɒⅿ” gʌ�”

Upload: phamnhu

Post on 26-Aug-2019

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Print-to-sound not print-to-meaning training helps ... · 24 adults each learn to read two di˝erent arti˚cial orthographies, each consisting of 24 novel words Manipulate focus of

Plaut, D. C., McClelland, J. L., Seidenberg, M. S., & Patterson, K. (1996). Understanding normal and impaired word reading: Computational principles in quasi-regular domains.Psychological Review, 103(1), 56-115. Taylor, J. S. H., Plunkett, K., & Nation, K. (2011). The in�uence of consistency, frequency, and semantics on learning to read: An arti�cial orthography paradigm. Journal of Experimental Psychology: Learning Memory and Cognition, 37(1), 60-76. Taylor, J. S. H., Rastle, K., & Davis, M. H. (2013). Can cognitive models explain brain activation during word and pseudoword reading? A meta-analysis of 36neuroimaging studies.Psychological Bulletin, 139(4), 766-791.Taylor, J. S. H., Rastle, K., & Davis, M. H. (2014). Distinct Neural Specializations for Learning to Read Words and Name Objects. Journal of Cognitive Neuroscience, 26(9), 2128-2154.

Print-to-sound not print-to-meaning training helps decoding and comprehension:Orthographic learning and fMRI

J. S. H. Taylor1, Matthew H. Davis2, and Kathleen Rastle1 1Royal Holloway, University of London, UK2MRC Cognition and Brain Sciences Unit, Cambridge, UK

1. Background

2. Method

5. Asymmetric differences in brain activity

6. Conclusions

7. References

3. Learning to read aloud and comprehend

4. Print–sound training benefits comprehension

Grant Number: ES/L002264/1

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

“fɒⅿ”See Say

See Say “camel”abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9

Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9

Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9

Day 1 Day 2 Day 3 Day 4 Day 6 Day 7 Day 8 Day 9

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1 2 3 4 5 7 86Number of times attempted

Number of times attempted1 2 3 4 5 7 86

Day 1

Day 2

Day 3

Day 1

Day 2

Day 3Day 4

Day 6 Day 7 Day 8 Day 9

Day 1

Day 2

Day 3Day 4

Day 6 Day 7 Day 8 Day 9

Day 1

Day 2Day 3

Orthography-to-Phonology Focus Orthography-to-Semantic Focus

00.10.20.30.40.50.60.70.80.9

1

0

1000

2000

3000

4000

5000

Reading Aloud:

Prop

ortio

n Ite

ms

Corr

ect (

+/- 1

SE)

Resp

onse

Tim

e in

ms

(+/-

1SE

)

Transfer Questions (Box 4)Does print−sound training improve comprehension?Does print−meaning traning improve reading aloud?

phonology

semantics

orthography

print-to-sound-to-meaning pathwayprint-to-meaning pathway

Models of reading propose two ways to comprehend written textPlaut et al., 1996

Corroborated by brain imaging dataTaylor, Rastle, & Davis, 2013

L intraparietal cortex

L posterioroccipitotemporal cortex

L inferior frontalgyrus

L anterior fusiform/temporal cortex

Simple Questions (Box 3)Does print−sound training improve reading aloud?Does print−meaning traning improve comprehension?

fаɪⅿ

fɒⅿ bᴂv ꜱⅰ:f

Orthography-to-Phonology (O−P) focus language

Orthography-to-Semantics (O−S) focus language

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄ abcedfghijklm-

nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-

nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-

nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

Print Sound

Meaning

24 adults each learn to read two di�erent arti�cial orthographies, each consisting of 24 novel words

Manipulate focus of training within subjects:

Orthography-to-Phonology focus: for one orthography, more training on systematic print−sound mappings

Orthography-to-Semantics focus: for other orthography, more training on arbitrary print−meaning mappings

Reading aloud:

Spelling:

Rhyme Judgement:

Comprehension:

Orthographic Choice:

Semantic Categorisation:

00.10.20.30.40.50.60.70.80.9

1

0

1000

2000

3000

4000

5000

Comprehension:

Trained Untrained Trained Untrained

Generalisation:Day 5 Day 10

Orthography-to-Phonology focusedtraining bene�ts accuracy and speedof reading aloud.

Orthography-to-Phonology Focus Orthography-to-Semantic Focus

If we equate practice on the speci�c task, does learning transfer to the alternative mapping?

Orthography-to-Semantic training does not transfer:Days 1-4, 6-9 also involved 3 x print−meaning tasks.This has no bene�ts for reading aloud.Instead, reading aloud accuracy depends onlyon practice with print−sound mappings.

Within regions active for [pseudowords − words] in EnglishImplicated in print-to-sound mapping

[Orthography−Semantic > Orthography−Phonology Focus]

Di�erences in activity during reading aloud

p < .001 uncorrected, p < .05 FWE cluster corrected

No di�erence in activity when saying word meanings

x = -50

x = -40

x = -20

phonology

semantics

orthography

Participants received pre-training in phonology−semantic mappings (establish an oral vocabulary).

Following this, we found that:Orthography−Phonology training was bene�cial for both reading aloud and comprehension.

Orthography−Semantic training was a slower way to learn to comprehend and was detrimental forreading aloud as well as generalising to untrained words.

Thus, even with minimal oral vocabulary training, it was easier for learners to use print−sound knowledgeto access word meanings, than to map from print-to-meaning directly.

Our precisely controlled experiment therefore supports existing UK teaching practice, which focuses onphonics (sounds and letters) in the early years.

Asymmetric bene�ts of print−sound and print−meaning training

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

x = -50

x = -20

x = -40

No significant differences in activity

Reading aloud is more e�ortfulfollowing orthography−semanticthan orthography−phonology training

posterior occipitotemporalcortex

precentral gyrus

intraparietal sulcus

angular gyrus

anterior middle temporal gyrus

superior frontal gyrus

No di�erence in e�ort of sayingmeanings following the two typesof training

“fɒⅿ”

0.7

0.75

0.8

0.85

0.9

0.95

1

1000

1500

2000

2500

3000

3500

Day 5 Day 10

See trained word Say pronunciation

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

Say meaning “camel”

Prop

ortio

n Ite

ms

Corr

ect (

+/- 1

SE)

Resp

onse

Tim

e in

ms

(+/-

1SE

)

Prop

ortio

n Ite

ms

Corr

ect (

+/- 1

SE)

Resp

onse

Tim

e in

ms

(+/-

1SE

)

See untrained word Say pronunciation “ⅿɒv”

See trained word

Trained Untrained Trained Untrained

Orthography-to-Semantic focusedtraining bene�ts speed but notaccuracy of saying meanings.

Orthography-to-Phonology focusedtraining bene�ts accuracy and speedof generalisation to untrained words.

Prop

ortio

n Ite

ms

Corr

ect (

+/- 1

SE)

Prop

ortio

n Ite

ms

Corr

ect (

+/- 1

SE)

Orthography-to-Phonology training transfers: Days 1-4, 6-9 also involved 3 x print−sound tasks.This transfers and improves saying the meaning.So comprehension accuracy depends more onprint−sound than print−meaning practice.

Comprehension: abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

Say meaning “camel”See trained word

Reading Aloud: “fɒⅿ”See trained word Say pronunciation

Day 0Sound−Meaning

pre-exposure

Days 1-4 and 6-9Print-to-sound tasks

O−P focus language x3 per dayO−S focus language x1 per day

Print-to-meaning tasksO−S focus language x3 per day

O−P focus language x1 per day

Day 11: fMRI scanReading aloud

Saying meaningsEnglish word reading

See Say

See Say “camel”

Hear Select - -

Hear Select

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄSee Selectabcedfghijklm-

nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

from all 24

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-

nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄabcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄ

abcedfghijklm-nopqrstuvwxyzÄ

See Select odd-one-outvehicle animal animal

Say “axe”Hear

See Say

Read aloud English words and pseudowords

Sparse imaging design - silent intervals between scans

[Orthography−Semantic vs. Orthography−Phonology Focus]

Training procedure Task Details

x3 per day

x1 per day pre-train

Print Sound

Meaning

x1 per day

x3 per day pre-train

O−P: O−S: See Say

O−S: O−P: See Say“tractor”

Within regions active for [words - pseudowords] in EnglishImplicated in print-to-meaning mapping

Days 5 and 10Test Sessions

ḃəʊꜱ �еꜱ gʌ�

vū:z

“gʌ�”

“fаɪⅿ”

“fɒⅿ”“fɒⅿ”

“lɒⅿ”

“gʌ�”