Title: Evaluation of the listening to speech programme: Developing speech discrimination skills in young ch
1Evaluation of the listening to speech programme
Developing speech discrimination skills in young
children
- Maggie Vance
- Dept of Human Communication Science
- University College London
2Speech discrimination language
- Speech perception skills in infancy correlate
with language development at 2yrs. (Tsao et el
04) - Auditory discrimination poorer in infants with FH
of SLI, and predictive of language at 3yrs. - (Benasich et al 02)
3Speech Discrimination difficulties
- Found in at least some children with
- Language Impairment Zeigler et al 2005
- Dyslexia Ramus et al 2003
- Significant History of Fluctuating Hearing Loss
Petinou et al 2001 - Auditory Processing Deficit Chermak
Musiek1997 - Speech Disorders Rvachew et al 2004
4Links with Literacy
- Foundation stage profile asks if child can hear
initial and final sounds in words. - Ability to discriminate phonemes a pre-cursor to
phonological awareness. - Difficulties not systematically addressed in
school intervention or in SLT
5Listening to Speech Programme Aims
- Develop a school-based resource
- For ALL children with less well-developed skills
- Promote speech discrimination skills
- to facilitate the development of
- Phonological awareness skills
- Understanding of spoken language
- New vocabulary and other aspects of language
6Delivery
- Children in Reception Classes
- Small Groups
- 30 minute sessions
- 3 x a week
- 6 weeks
7Minimal Pair Picture Tasks
- Range of activities
- Hierarchy of difficulty of contrasts
- sheep / ship
- school / stool
- Some practice against background noise
- Some practice with multiple speakers
8Example
- Fridge magnets
- Put picture in back of lorry
- Picture lotto
- Draw round the picture
9Participants
- 10 primary schools in SE England
- 308 children in Reception classes assessed
- Speech Input Processing in Children
- (Vance, Rosen, Coleman)
- Linguistic Concepts sub-test CELF-P
- 56 selected for intervention
- The 4 or 8 children in each school with weakest
profiles on assessment
10Speech Input Processing in Children (SIPc)
XAB task
11Whole Group Findings
- Children aged 4-5 years
- SIPc
- Linguistic concepts on CELF-P
- RAPT
- Significant correlations between speech
discrimination, and verbal comprehension
expression
12Intervention Group Characteristics
13DesignCluster randomisation of schools Blind
reassessment at T2 and T3 (mostly)
14Intervention Groups Matched
15(No Transcript)
16Outcomes
- MAIN OUTCOME MEASURES
- Speech Discrimination SIPc
- Receptive language CELF-P
- SECONDARY OUTCOME MEASURES
- Expressive language - RAPT
- Teacher Questionnaire - attentive listening
- Digit span WMTB-C
- Rhyme awareness PIPA
- POSSIBLE PREDICTIVE VARIABLES
- Non-Verbal abilities - Ravens Matrices
- Hearing History Questionnaire
- Hyperactivity Conduct ratings SDQ
- Speech output EAT
17- Intervention vs No Intervention
- No significant difference between intervention
groups on any outcome measure - both improved
equally
18- Intervention vs No Intervention
- No significant difference between intervention
groups on any outcome measure - both improved
equally
19Age and Experience
- Are improvements in speech discrimination at T2
- to do with age?
- Probably not (no sig correlation with age)
- to do with school experience?
- Dont know?
- to do with test experience?
- Probably!
20At T1 Group 1 11/28 not able to do SIPc
Group 2 7/28 not able to do SIPc
21Rhyme Awareness (PIPA)
22Group 2 comparing T1-T2, T2-T3
23Predictors and base-line scores
- Baseline scores in receptive language (CELF-P)
significantly predicted by - Baseline teacher questionnaire
- Baseline scores in speech discrimination (SIPc)
significantly predicted by - Hearing History questionnaire
- Conduct rating
- BUT not with
- non-verbal skills speech production
hyperactivity rating.
24Predictive Variables
- Increases in receptive language (CELF-P) (T1-T2)
significantly predicted by - Hearing History Questionnaire
- Increases in speech discrimination (SIPc)
significantly predicted by - Hearing History Questionnaire
- Baseline teacher questionnaire
25Predictive Variables
- Final Score in receptive language (CELF-P)
significantly predicted by - Hyperactivity rating
- Final Score in speech discrimination (SIPc)
significantly predicted by - Baseline teacher questionnaire
- Non-verbal skills
26Pitfalls
- Group
- Heterogeneous group
- selection criteria too broad
- Subtle differences between Intervention Groups
- Measurement
- Outcome measurement
- Not sufficiently robust (SIPc) (Noise not done at
T2) - Not sufficiently sensitive (CELF-P)
- Practice effects on baseline / outcome assessments
27- Programme
- Lack of developmental profile of discrimination
of contrasts for planning programme - Generic programme not tailored to individual
childrens profiles
28Thanks to
- The Health Foundation
- Stuart Rosen and Mike Coleman
- DHCS students