Post on 09-Feb-2022
The
Sound Training StudyPrinciple Investigators: Genevieve McArthur, Max Coltheart
Research Support: Danielle Ellis, Carmen Atkinson, Sarah Barry, Eliza Fraser
Technical Support: Craig Richardson, Rob Seiler, Noel Stephenson, Steve Saunders
Administrative Support: Suzanne Mendes, Katie Webb
Moral support: Romina Palermo, Steve Mondy, Martha Turner
Child support: Lyndsey Nickels, Sachicko Kinoshita, Karen Smith-Locke
External support
• CHW• Dr Pam Joy• Dr Natalie Silove
• CEO• Frederick Patchell• All special ed teachers
• Therapists• John Dyson • Jeanie Raffills• Joni Coleman• Trish Hurley• Samantha Barros• Jennie McLean
• Schools• Booragul Primary
School• St Catherine’s School
Waverley• Holy Spirit Primary
School• Mowbray Public School• Nepean District
Christian School• Special Education
Centre at Macquarie University
• All children and families
Introduction
• Specific language impairment (SLI)• unexplained failure to acquire spoken
language skills • grammar, phonology, semantics• ~ 5%
• Developmental dyslexia• unexplained failure to learn reading skills• ~ 5%
What causes SLI and dyslexia?• Heterogeneous conditions• No single cause• Multiple risk factors
• limited processing (SLI)• poor grammatical rule learning (SLI)• poor phonological short-term memory (SLI)• poor phonological processing (SLI and dyslexia)• poor automatisation (SLI and dyslexia)• poor visual processing (dyslexia)• low-level sound processing deficit (SLI and dyslexia)
How?Less able to process sounds
Less able the process speech sounds
Poor phonological representations
Poor receptive and expressive phonology
SLI
Poor phonological decoding
Dyslexia
What type?
• Rapid auditory processing• SLI: Tallal & Piercy (1973a)• Dyslexia: Tallal (1980)
• Frequency discrimination• SLI: McArthur & Bishop (2004)• Dyslexia: Baldeweg et al. (1999)
• Speech processing• Studdert-Kennedy & Mody (1995)
• Association does not mean causation
Does poor sound processing cause SLI or dyslexia?
• Tallal, Merzenich et al. (1996)• Training programme
• Frequency-modulated tones
• CV syllables• CVC syllables• Words• Phrases• Sentences
• 11 children SLI• Lab: 3 hours per day, 5 days a
week, for 1 month• Home: 1-2 hours per day, 7
days a week, for 1 month• Significant gains in spoken
language tests
• Claim: “The measured improvement in a child’s threshold for correctly segmenting and sequencing successive nonverbal auditory stimuli was significantly correlated with post-training outcome in on-line language processing”
• !!!!! Correlation does not mean causation!!!!
Fast ForWord• Claim: “Once a programme is found to be
effective, it is a reiterative process to determine what specific training components are most effective, so that the programme can be improved. It is anticipated that many future studies will address these issues.” (Tallal, 2000, p. 151).
• Where are they?• Does training low-level sound processing
improve spoken language or reading skills?
Questions• How many children with SLI or dyslexia
have a sound processing deficit?• What type do they have?• Can sound processing deficits be fixed?• Does this improve spoken language,
reading, spelling, or phonological processing?
Children Dyslexia SLI Control
N 66 28 38
Age (months) 82-183* 97-168* 86-148*
Non-verbal IQ (ss) 80-131* 80-126* 89-131**
Recalling Sentences (ss) 4-13** 1-13* 7-16***
Nonword Repetition (ss) 6-14** 3-13* 8-15***
TROG (ss) 60-120* 55-106**
BPVS (ss) 83-121* 68-108** CC Nonwords (z) -2.41-2.37* -2.41-0.01* -0.94-2.41**
CC Irregular words (z) -2.41-1.32* -2.41-0.57* -0.99-1.94** no reported sensory, neurological, physical, emotional, social problems
Good Bad
Pre-training tests
• Spoken language• Recalling sentences (Clinical Evaluation Language
Fundamentals -4)• Nonword repetition (NEPSY)
• Reading• Nonwords (Castles and Coltheart lists)• Irregular Words (Castles and Coltheart lists)
• Spelling• Regular words (Test of Written Spelling)• Irregular words (Test of Written Spelling)
• Phoneme Discrimination• Alliteration test with pictures (Phonological
Assessment Battery)
Pre-training tests
• Sound processing tests• Frequency discrimination• Rapid auditory processing• Vowel discrimination• Consonant-vowel discrimination
• Threshold scores• If produced abnormal score on first run then
did again to check that score was valid
• Target impaired sound processing• Frequency Discrimination• Rapid auditory processing• Vowels• Consonant vowels
• Home training• Laptop• ½ hour per day• 4 days per week• 6 weeks
Training
Post-training tests• Spoken language
• Recalling sentences (Clinical Evaluation Language Fundamentals -4)
• Nonword repetition (NEPSY)• Reading
• Nonwords (Castles and Coltheart lists)
• Irregular Words (Castles and Coltheart lists)
• Spelling• Regular words (Test of Written
Spelling)• Irregular words (Test of Written
Spelling)• Phonological awareness
• Alliteration test with pictures (Phonological Assessment Battery)
• Sound processing tests• Frequency discrimination• Rapid auditory processing• Vowel discrimination• Consonant-vowel
discrimination
Delayed tests• Spoken language
• Recalling sentences (Clinical Evaluation Language Fundamentals -4)
• Nonword repetition (NEPSY)• Reading
• Nonwords (Castles and Coltheart lists)
• Irregular Words (Castles and Coltheart lists)
• Spelling• Regular words (Test of Written
Spelling)• Irregular words (Test of Written
Spelling)• Phonological awareness
• Alliteration test with pictures (Phonological Assessment Battery)
• Sound processing tests• Frequency discrimination• Rapid auditory processing• Vowel discrimination• Consonant-vowel
discrimination
-2
0
2
4
6
8
10
12A
vera
ge S
ound
Pro
cess
ing
Thre
shol
d (t
)
How many children with SLI or dyslexia have a sound processing deficit?
Dyslexia SLI Control
F(2, 131) = 9.69, p < .001)
individual’s mean threshold for all 4 tasks
22/66 14/2833% 50%
Limit of normal scores
-2
-1
0
1
2
3
Rap
id A
udito
ry P
roce
ssin
g (t
)
Rapid Auditory Processing
0
10
20
30
Fre
quen
cy D
iscr
imin
atio
n (t
)
Frequency Discrimination
What type?
Dyslexia SLI Control22/66 9/2833% 32%
Dyslexia SLI Control4/66 1/286% 4%
F(2, 131) = 6.49, p = .002)
F(2, 130) = 5.92, p = .003)
-2
0
2
4
6
8
Co
nson
ant-v
owel
Dis
crim
inat
ion
(t)
Consonant-vowel Discrimination
-2
0
2
4
6
Vow
el D
iscr
imin
atio
n (t
)
Vowel Discrimination
What type?
Dyslexia SLI Control11/66 8/2817% 29%
Dyslexia SLI Control12/66 8/2818% 29 %
F(2, 129) = 3.26, p = .04) F(2, 131) = 2.82, p = .02)
FD RAP Vowel CV FD RAP Vowel CV
Can frequency discrimination be trained?
1: Mean improvement from pre-tests to immediate post-tests for FD trainees
No training effect = 00
5
10
15
Impr
ovem
ent i
n Fr
eque
ncy
Dis
crim
inat
ion
(t)Mean improvement from pre- to post-testing
95% CI; if falls over 0 then no sig change
SLI/dyslexia FD trained Untrained controlsN = 19 N = 38
Frequency Discrimination
Can rapid auditory processing be trained?
-10
-5
0
5
10
15
Impr
ovem
ent i
n R
apid
Aud
itory
Pro
cess
ing
(t)
FD RAP Vowel CV FD RAP Vowel CV
SLI/dyslexia RAP train Untrained controlsN = 3 N = 38
Rapid Auditory Processing
Can vowel discrimination be trained?
-20
-10
0
10
20
30
40
Impr
ovem
ent i
n Vo
wel
Dis
crim
inat
ion
(t)
FD RAP Vowel CV FD RAP Vowel CV
SLI/dyslexia V trained Untrained controlsN = 3 N = 38
Vowel Discrimination
Can consonant-vowel discrimination be trained?
-2
0
2
4
6
8
Impr
ovem
ent i
n C
V D
iscr
imin
atio
n (t)
FD RAP Vowel CV FD RAP Vowel CV
SLI/dyslexia CV trained Untrained controlsN = 10 N = 38
Consonant-vowel Discrimination
Does this improve spoken language?
Recalling Sentences Nonword repetition
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
Impr
ovem
ent i
n R
ecal
ling
Sent
ence
s (s
s)
-2
-1
0
1
2
Impr
ovem
ent i
n N
onw
ord
Rep
etiti
on (s
s)
Imm Del Imm Del Imm
Dyslexia SLI Controls
Imm Del Imm Del Imm
Dyslexia SLI Controls
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
Imp
rove
men
t in
Pho
nolo
gica
l Dec
odi
ng (s
s)
Does this improve reading?
Phonological Decoding Sight word reading
-0.3
-0.2
-0.1
0.0
0.1
0.2
0.3
0.4
Impr
ovem
ents
in Ir
regu
lar W
ord
Rea
ding
(ss)
Imm Del Imm Del Imm
Dyslexia SLI Controls
Imm Del Imm Del Imm
Dyslexia SLI Controls
Does this improve spelling?
Regular words Irregular words
-2
0
2
4
Impr
ovem
ent i
n R
egul
ar W
ord
Spel
ling
(ss)
-7.5
-5.0
-2.5
0.0
2.5
5.0
7.5
Impr
ovem
ent i
n Irr
egul
ar W
ord
Spel
ling
(ss)
Imm Del Imm Del Imm
Dyslexia SLI Controls
Imm Del Imm Del Imm
Dyslexia SLI Controls
Does this improve phoneme discrimination?
-5
0
5
10
15Im
prov
emen
t in
Phon
eme
Dis
crim
inat
ion
(ss)
Imm Del Imm Del ImmDyslexia SLI Controls
Answers• How many children have a sound processing
deficit?• Dyslexia: 33%• SLI: 50%
• What type do they have?• FD>Vowel=CV>RAP
• Can sound processing deficits be fixed?• Yes
• Does this improve spoken language, reading, spelling, or phonological processing?• No
Fast ForWord• “Once a programme is found to be effective, it is a
reiterative process to determine what specific training components are most effective, so that the programme can be improved. It is anticipated that many future studies will address these issues.” (Tallal, 2000, p. 151).
• Fast ForWord• Frequency-modulated tones • CV syllables• CVC syllables• Words• Phrases• Sentences
• 2 hours per day, 5 days a week, for 6 to 8 weeks• Focus training on tasks that might work (good for FFW)• Halve the amount of training (good for kids)
Confirmation
• We are all on the same page• Research so far supports current speech,
language, and reading practice• Hit the problem at the level of the problem:
spoken language and reading
These kids can learn
• But they need much more practice• Not enough time, money, or therapists• Children can be persuaded to practice with
1000s of boring sounds using software• What could be done if you could use
software to get them to practice with 1000s of trials of relevant spoken and written verbal stimuli?