home index by author index by subject

There is a PowerPoint file available to accompany this presentation. The BDA Conference 2001 CD-ROM contains 61 PowerPoint files. For details of how to obtain the CD-ROM, please contact the BDA.

Keynote

Friday stream K Session 09.00 - 09.45 Length 25 minutes

Individual differences in dyslexia: The Dyslexia Spectrum

Margaret Snowling

University of York Mjs19@york.ac.uk

Abstract

There is now a large body of evidence showing that, in cognitive terms, dyslexia can be characterised as a core phonological deficit. The phonological deficit theory of dyslexia has been highly successful because it makes contact with findings from the normal development of reading and it provides a parsimonious explanation of dyslexia across the life-span, as well as offering a principled approach to remediation. The theory is not without its critics however. First, it is not clear how the phonological deficit hypothesis can account for the individual variation seen among dyslexic readers. Second, because of the reciprocal interaction of reading and phonological skills , it is possible that the phonological deficits in dyslexia might be exacerbated by literacy failure (and less obvious in transparent than opaque orthographies). Moreover, the relationship between dyslexia and other learning disorders is not fully understood. This paper will provide an overview of recent studies in support of the phonological deficit hypothesis. Drawing on evidence for studies comparing dyslexic children and children with oral language impairments, and from a longitudinal study of children at genetic risk of dyslexia, it will propose that the heterogeneity of dyslexic samples can be explained by differences in the severity of phonological impairments, and differences in language skills that 'bootstrap' reading development. The clinical utility of a spectrum of dyslexic difficulty is discussed.

Dyslexia and Cognition

"There is virtually unassailable evidence that poor readers as a group are impaired in a very wide range of cognitive tasks in the phonological domain"
- Share, 1995 p186

Dyslexic impairments in adulthood

Dyslexia and Behaviour
Phonological representations and literacy development

Development of phonological representations?

• Normal development
  - Lexical development (vocabulary)
  - Learning (letter knowledge)
  > Carroll & Snowling in press
• Processing mechanisms
  > Input (perceptual) processes
  > Output (production) processes

Experiment 1

Input vs output processing

• 12 year-old dyslexic readers, CA- and RA-controls
  - Rapid auditory perception task (after Tallal)
  - Nonword repetition task
  - Explicit phonological awareness task
  > Marshall, Snowling & Bailey, (in press)

Tallal RAP task

• Child learns to associate two tones differing in pitch with two responses
• Child hears two tones and repeats the sequence - high - low; low-high; high-high; low-low
• The time interval between the tones is varied from long (150ms) to short (10 ms)

Marshall, Snowling & Bailey (2001)

• o Nonword repetition
• o Phoneme deletion
  > "bice" - [b] = ?
  > "cleaf" -[k] = ?

Summary

• A small group of dyslexic readers have difficulty on the RAP task
• Difficulties in output processes are more pervasive
  - the dyslexics who did poorly on the RAP task were NOT those that had more severe phonological deficits
  - Similar findings from speech perception task

Experiment 2

Spoken word recognition vs retrieval

o Input How much auditory input is required to recognise a spoken word? - Gating o Output How efficiently can a name be retrieved for a familiar word? - Rapid Naming Griffiths & Snowling (2001)

Forward gating procedure

Participants

Time to identify the words

Summary

• Processing of auditory input to identify spoken words is normal in dyslexia
• Retrieval of input for naming tasks (animals, furniture, colours) is impaired

Experiment 3

Recognition vs recall

• Memory scanning
• On each trial subject hears a list of between 1 and 4 items
• Pause, then a probe item presented
• Subject responds (Y/N) by key press - RT measured
• Memory span assessed for 1 & 5 syllable words and 1 & 3 syllable nonwords
  > Weighall, Hulme, Griffiths & Snowling (in preparation)

Participants

• 15 Adult dyslexic readers
  - Mean IQ 120, WRAT Spelling SS 49-99.
• 24 student participants
  - comparison group Hulme et al. (1998)

Memory span for each group

Memory Scanning

Summary; memory study

• Span
  - Dyslexics have shorter spans, especially for long words
• Scanning
  - Dyslexics slower but no less accurate than controls

Implies;

• o contents of memory representation normal
• o output effects
  - preparatory intervals, item duration and longer pauses before long words

Deficits in phonological representation

o Group deficit in dyslexia

- Input processes
> Normal
- Output processes
> Impaired

Why do dyslexics differ in reading behaviour?

• Differ in severity of phonological deficit
• Differ in pervasiveness of phonological deficit
  - Input / output mechanisms
  - development: delay versus disorder
• Differ in reading practice

Assessment battery

Griffiths & Snowling (2001)
59 12-yr dyslexic readers
59 RA-controls (8-yrs)
- Nonword reading
> chob, foop, tadlen
- Exception word reading
> ocean, yacht, sergeant

Predictors of reading profile

• Phonological skills
  - Phoneme deletion
  - Rhyme production
  - Nonword repetition
• Verbal STM
  - Memory span
  - Speech rate
• Print exposure
  - a measure of how much children read

Have phonological deficits in dyslexia led us astray?

Family study of dyslexia

• 74 children at genetic risk of dyslexia
• 37 controls from families with no history of dyslexia
  - seen at 3;09, 6 and 8 years
• WORD composite at 8 years
  - 66% at risk impaired
  - 14% controls
• Gallagher, Frith & Snowling (2000)

Summary

What is the relationship between language and literacy?

• Path analysis
  - Pre-school (4 yrs)
• Oral Language
• Letter knowledge
  - Age 6
• Vocabulary
• Phonological awareness
• Predict phase 3 (8 yrs)
  - Basic reading skills (word level literacy)
  - Reading comprehension Comprehension

Pathways to literacy

• Pre-school language and letter knowledge underpin phonological awareness at 6 years
• Phonological awareness and phonic skills at 6 years are unique predictors of word-level literacy at 8
• Pre-school language predicts vocabulary at 6 yrs and later reading comprehension

Broader phenotype?

• Poor letter knowledge at 4 yrs (learning impaired)
• Deficient use of phonics (OP mappings) at 6 yrs
• Reading development at 8 yrs bootstrapped by stronger speech/language resources
  - weak spelling
  - fail later?

Triangle model

Plaut, Seidenberg and colleagues

Deficit and Compensation

• Division of labour between semantic and phonological pathway
  - Plaut et al. 1996
• If development of OP mappings constrained, rely on language skills outside of phonological module
  - Nation & Snowling, 1998

Individual differences in dyslexia

• o Nature and severity of deficit in phonological representations
  - delayed development (poor vocabulary)
  - deficient processes (output)
• o Integrity of language skills outside phonological module
  - protective factor
  - compensatory mechanisms

Clinical implications 1

• Reading development best conceptualised as an interactive process, the demands of which change with time
• Children with deficits in phonology OR language are at risk of dyslexia

Clinical implications 2

• Utility of dimensional view
• Phonological deficits at core, modified by other language skills
• Concept of "dyslexia-spectrum" disorders

Conclusions 1

• Phonological deficits at core of dyslexia
• Individual differences in dyslexia depend upon the severity of the phonological processing deficit
• Modified by cognitive and language skills that may function as compensatory factors over the course of reading development

Conclusions 2

• There is continuous variation in reading at the behavioural level [rather than pure sub-types]
• IDs are determined by an interaction of
  - Phonological skills
  - Semantic (and syntactic) skills
  - Teaching and development