Screening and intervention: a stitch in time saves nine!

1. Screening and intervention: a stitch in time saves nine! Professor Angela Fawcett Centre for Child Research University of Swansea Danish Neuropsychological Society February 2009

2. Plan of Talk • Evidence for early screening • Towards screening tests • The Dyslexia Screening Tests (3.5-75!) • Linking Intervention to screening • The Pre school Screening Test • Conclusions – the way forward! Acknowledgements: Rod Nicolson Psychological Corporation, Leverhulme Trust, Nuffield Foundation

3. Key Questions for Dyslexia Research • Q1. Can we identify dyslexia and other learning disabilities before a child fails to learn to read? • Q2. Can we help these children to catch up with their peers and make this cost-effective? • Q3. Can we identify those children who will be difficult to accelerate? • Q4. Do we need different methods to teach dyslexic children? If so, what? • Q5. What should we do next?

4. Why screen for dyslexia? • Commonest developmental disorder (around 5%) • Genetic origin, 50% chance of child suffering if their parent is dyslexic, persists into adult life • Appropriate teaching support improves reading, phonological skills, spelling and confidence (Rack, 1994) • Lack of support leads to continuing failure, problems become entrenched and generalised • The earlier support can be provided, the more cost-effective it will be. The stitch in time approach • Our aim - to identify problems before children fail!

5. US interventions (Torgesen, 2001) • 10 year programmes funded by Govt / NICHD (~$1billion) • i) ‘Pure’ phonological interventions (Method A vs B) do not generalise to reading - disappointing! Change in emphasis to add fluency (cf. Wolf) • ii) ‘Closing the gap’ at secondary level. Special ed improves reading only 0.04 st. dev (class sizes!) • iii) ‘Junior’ level (8-11) Need individual intensive intervention, 67.5 hrs Gains per hour of instruction 0.2 ss • iv) Prevention more effective and cost effective than remediation!

6. Phonics intervention is most effective for reading in 1st grade Data from the National Reading Panel (0.2, 0.5, 0.8 small, mod and large ES)

7. Towards Screening Tests for Dyslexia • Requirements • Theoretically Defensible • Valid, objective, reliable, normed • Practically Useable • quick, simple, intelligible • Politically acceptable • Parents, Children, Schools, Teachers, Local Authorities, Educational Psychologists, Government • the tensions between these various requirements mean that any screening test is a compromise between thoroughness and cost!

8. Screening, Support, PolicyThe blue-print

9. Approach adopted in developing tests • Develop a 10 sub-test, 30 minute diagnostic battery of theoretically motivated tests covering a wide range of skills, cognitive, literacy, phonological, speed, knowledge and motor • Must be quick and easy to administer, in pencil and paper format • Gather performance data from over 1000 ‘normal’ children for the DEST, 800 children and 60 dyslexic children for the DST, and 1100 adults for the DAST • Analyse the norm data and adjust the instructions and content, based on feedback from users • Construct norms for the DEST (4:6 to 6:5 yrs) the DST (6:6-16:5 yrs), and the DAST (16:6 to 75!). • Publish with the Psychological Corporation 1996/8 - now best sellers and 2nd editions published!

10. The Dyslexia Early Screening Test The DST and DAST measure fluency, including 1 minute reading and copying, nonsense passage, spelling, verbal and semantic fluency • .

11. DEST (4:6-6:5) Cognitive/phonological/speed Test 1 Rapid Naming Test 3 Phon’ Discrim Test 5 Rhyme detection Test 6 Digit Span Test 7 Digits Test 8 Letters Test 9 Sound order Motor Test 2 Beads Test 4 Postural Stability Test 10 Shape Copying DST (6:6-16:5) Cognitive/phonological/speed Test 1 Rapid Naming Test 3 One minute reading Test 5 Phon’ Segmentation Test 6 Two minute spelling Test 7 Backwards Digit Span Test 8 Nonsense passage read Test 9 One minute writing Test 10 Verbal fluency (S) Test 11 Semantic fluency Motor Test 2 Beads Test 4 Postural Stability Components of DEST and DST tests

12. Phonological skill:DEST, DST, DAST • A series of phonological tests designed to tap more complex skills as the child develops into an adult • I )a) Rhyme and b) Discrimination DEST • A) Can the child say whether or not two words rhyme? • leg hen pin win cup pup hop doll toe toad • B) Can the child say whether or not two words are the same ? • bad, dad cap, cap. Turn away so child can’t lip read • 2) Segmentation DST/DAST • Say stake Say it again, but without the /st/ • Say stake Say it again, but this time without the /s/ • 3) Spoonerisms DAST/DST secondary (in press) • Sean Connery • Shirley Bassey

13. DEST, DST, DAST : Rapid Automatised Naming

14. DEST Test 9: Sound order • Child listens to pairs of sounds (a squeak [mouse] and a quack [duck]) on tape, and has to say which one was first, the mouse or the duck. The separation between the sounds is steadily decreased, making the task harder and harder. • This is based on work by Paula Tallal which indicates that language-disordered children often have difficulties with this type of rapid temporal processing task.

15. DEST: Knowledge and working memory DEST Test 7: Digit Naming • The child has to say the names of the following digits. A straightforward test of knowledge. • 4, 8, 3, 9, 5, 7, 6 DEST Test 8: Letter Naming • The child has to say the names of the following letters • t, s, d, e, w, o, b, q, n, y DEST Test 6: Digitspan. DST Test 7: Backwards Digitspan • Standard tests of memory for a series of numbers

16. DEST Test 10: Shape Copying DEST and DST Bead threading. How many beads can you thread in 30 seconds?

17. DEST, DST and DAST: Postural Stability • Balance Tester • Not to scale! • How much the child wobbles when pushed gently in the back. • Standard test of cerebellar function (balance) score 0-6

18. DST and DAST: Fluency in writing • This is designed to assess speed of writing. Dyslexic children are slower (and less neat) writers. Writing speed is currently one of the key issues for examination allowances. Score is primarily the number of words transcribed in one minute, with adjustments made for errors. • I am copying a short passage to check my speed of writing. I have one minute to finish as much as I can. I should work quickly but accurately, so that my handwriting can be read.

19. DST and DAST 2 Minute Spelling

20. DST 1 minute reading (sample) Read down as fast as you can! Say ‘pass’ if you don’t know the word pat fog king oil golf tea fun gear Nonsense word passage sample Age 6:6- 8:5 One day a mib fell into the feg. He was so sup and tid that he cried. Age 12:6-16:5 In the olden days, a rennifer set out to craiberg an enormous dollitroy that threatened his country. Time and accuracy scored. Time bonus & penalties. Score 2 for nonsense words Fluency in reading (DST, DAST)

21. Non-verbal Reasoning (DAST) • Dyslexic adults have strengths in reasoning ability, especially if the reasoning involved is not language-based. • Examples • I) Find the next one in the sequence • 2) Anologies - A is to B as C is to ? • Included in DST-S (2004)

22. DST and DAST. Verbal & Semantic Fluency • Test 10 is a test of verbal fluency, and is simply how many words beginning with S the subject can think of in a minute. Test 11 is for semantic fluency – how many animals the child can think of in a minute. • Uta Frith and her colleagues have shown recently that dyslexic children may well score relatively poorly on verbal fluency, but relatively well on the semantic fluency task (test 11).

23. Scoring the DEST o - -- -- - - - o Each test score is converted into an age-appropriate normed score, using the norms we have gathered in out national testing programme. There are separate norms for 4:6 - 4:11, 5:0 - 5:5, 5:6 - 5:11, 6:0 - 6:5 -- -

24. Working out the ‘at risk’ score and profile • Fill in the raw score for each test on the score sheet. Compare scores with norms for profile. • 0 is average, - below, + above. DST has 3 - (0-4, 5-11, 12-22nd centile respectively) DEST has 2+ to pick out high achievers • The profile shows which areas to address for the IEP • Work out the ‘at risk’ score by scoring 1 mark for each - score, (0 for the rest) and dividing by 10. • An ARQ of 0.9 is strong risk, and 0.6 mild risk

25. Case study: the DEST • age: 5:5 • language delay, lethargic, head injury • Outcome • -- 3, - 5, total (2 x 3) + 5 =11 • At Risk Quotient 1.1

26. DEST Validation Study (100 children) • Tested on DEST at 5 • Re-tested on reading and spelling at 7 • Hit rate =90%, False positive Rate = 12% • Criteria 0.9 DEST, 10 month reading deficit • DST and DAST hit rate 94% dyslexics, 0% controls

27. Validation of DST with our Dyslexic Panel

28. Central Lancs study (Whiteley 2002 • Independent study using the DEST • 430 children screened • 90 ‘at risk’ compared with matched controls • On all DEST subtests very significant impairments for the at risk group • NB The postural stability test is not significant because it only identifies children with dyslexia, not those with general difficulties. I shall return to this when looking at intervention.

29. Support - General approach adopted • Screen for potential problems, using teacher recommendations and/or screening tests • Develop an IEP for each child on the basis of profile of strengths, weaknesses, attainments and objectives • Create groups of 2-3 children with similar IEP’s • Give structured group support in 2 half hours per week • Evaluate progress at the end of 10 weeks support

30. Study 1 Screening and Support (Infant) Nicolson, Fawcett, Nicolson, Moss and Reason, 1997 • Schools in deprived / non-deprived areas in 2 cities • 16 children identified 'at risk' in each school [64 children, age 6.0 (5.5-7.0), RA 5.6 (5.3-6.0)], plus Controls • Pretest. • WORD tests of reading and spelling, BPVS vocabulary, Dyslexia Early Screening Test (DEST) • Intervention • Interactive Assessment & Teaching intervention in groups of four, for 20-30 minute sessions, three times a week for 10 weeks • Post-test. • WORD reading and spelling

31. Study 1 Results of intervention Significant improvement for trained group, standard score for reading ‘Accelerating’ from 88.9 to 93.2. Controls deteriorated slightly

32. Study 2. Effectiveness of Junior Intervention Fawcett, Nicolson, Moss, Nicolson and Reason (2000) Schools selected in deprived / non-deprived area in 2 cities 16 children identified 'at risk' in each school [36 children, age 7.6, reading Standard Score 79.9, spelling 82.6]. 51 Control ‘at risk’ children • Pretest. WORD reading, spelling, BPVS vocabulary, DST • Intervention ‘IA & T’ intervention in groups of two, for 30 minute sessions, twice a week for 10 weeks • Post-tests WORD reading and spelling (curriculum-based reading skills) • Follow-up (6 months later) WORD reading and spelling

33. Results of Junior school evaluation • Significant improvement in reading and spelling (3.4, 5.7 respectively) persisted 6 months later. • Controls 1 and 2 point improvement (NB Literacy Hour!)

34. Central Lancs Intervention StudyWhiteley, H., Smith, C., Godwin, M., & Oakley, S. (2002). Linking early literacy screening and intervention: Recognising individual differences. Policy on Dyslexia Conference, Uppsala. • 12 schools visited, 432 children reception screened on DEST • 20.8% at risk • All at risk children: 12 week intervention 20 mins per day small groups ‘Launch into reading’ • 65 children. • 40 ‘beneficiaries’ [WRAT reading and spelling], 25 non-beneficiaries • Non-beneficiaries significantly worse on DEST only for auditory discrimination and postural stability (replicates our study) • Further 15 weeks individual support for non-beneficiaries • Overtook beneficiaries, but still 19 ‘non-beneficiaries’ • Policy implications • DEST valuable predictor both of risk (ARQ) and need (profile) • Screen early, consider individual differences

35. The Qinetiq study This study used the DEST in conjunction with specially designed glasses to evaluate beginning readers - in collaboration with UK government • Screening study: 241 children tested at age 5 Children at risk on DEST were significantly worse on Reading and Spelling. • DEST better than Teachers, Verbal IQ or background information. (Teachers identify younger children!) • Follow up data • 98 children followed over 6 months to see how effective DEST was in identifying risk. DEST only missed 1 poor reader! • Improvement in reading in DEST risk group only 1.2 words, vs 4.5 for the no risk group. • Without extra support, children at risk don’t keep pace with their peers!

36. Implications of poor performance on DEST

37. 2nd editions 2nd editions of DEST and DST were published in 2004/5, allowing us to improve on the 1st editions. DEST 2 New tests of visual memory and vocabulary • Two different versions of the DST have been published DST-J for children aged 6.5-11.5, DST-S for children aged 11.6-16.5 DST-J New tests of vocabulary, more sensitive 1MR, rhyming DST -S Non-verbal reasoning, spoonerisms (modified from DAST) Modifications based on continuing research and feedback from teachers Computer program provided to enter raw scores, generating profile and report

38. Conclusions on DEST, DST and intervention Small group intervention can be effective and cost-effective 1) DEST valuable predictive index of children likely to need specialist support. 2) Early intervention in a child’s first year at school (5/6 years), before literacy problems are compounded by the ‘Matthew effect’ is likely to be the most effective approach. 3) Junior school children (7+) with low ‘at risk’ scores on the Dyslexia Screening Test have a good chance of catching up with their peers following a relatively short intervention 4) However, those with high DEST and DST ‘at risk’ scores are likely to need continuing support over a relatively long period. • Implications for intervention – the earlier the better!

39. 6. The Pre-School Screening Test

40. PREST (3.5-4.5 years) • PREST involves two halves, each around 15 minutes. • PREST 2 is only given if possibly ‘at risk’ on PREST 1.

41. Visual Memory: Corsi frogs • A spatial memory game, jumping a frog on to the lily pads. • AJF’s fun version of Corsi’s traditional memory test which uses boxes.

42. PREST 2

43. PREST Test 10: Rhyming

44. Example of a PREST profile of needs

45. The Sheffield pre-school Intervention • 60 4 year old children screened with PREST in mixed socio-economic area • 20 'at risk' children identified - Intervention undertaken. • 12 control children (same nursery) received no intervention • Significant improvements in a number of key academic skills • Children 'accelerated' to have good chance of keeping up with their peers in infant school. • iii) 13 children with special needs including siblings of dyslexics (see Scarborough, 1991 for early deficits), LD, autism, deafness also had intervention. • iv) Non-improvers had 2nd phase 6 weeks (targeted 1 to 1) • Again results were very encouraging!

46. Effect sizes on Screening and intervention with the PREST Intervention group worse than controls at pre-test Overtook controls after intervention Digit span untrained!

47. Long term effects of interventionFawcett, Lee and Nicolson (2003) • Children were classified by risk, with 0.5 here indicating mild risk • Intervention group 85% risk at pre test, 0% at 5.8. • Special group 83% risk at pre test, 16% at 5.8. • Even children with known SEN accelerated!

48. Intervention at 4 and Literacy at 7 • In the UK all children take standardised tests known as SATS in the same week at age 7. • At 7 children should be at level 2 or above. • Scores are given as 2a, 2b, 2c, 2, 2d. coded as 2.75, 2.50, 2.25, 2.0 and 1.9 respectively • Children in the intervention group had higher mean scores on reading and maths than the control group (2.75 vs 2.42 and 2.92 vs 2.50 for reading and maths respectively) • Effect sizes were 0.87 for reading and 1.18 for maths A stitch in time saves nine!

49. Key Questions for Dyslexia Research • Q1. Can we identify dyslexia and other learning disabilities before a child fails to learn to read? Yes! • Q2. Can we help these children to catch up with their peers and make this cost-effective? Yes! • Q3. Can we identify those children who will be difficult to accelerate? Yes! • Q4. Do we need different methods to teach dyslexic children? If so, what? I believe so • Q5. What should we do next? More research

50. Conclusions – the way forward • We have constructed 30 minute screening tests, based on existing theoretical findings, which can be used from 3.5 years up. Ready to Learn is now published in the US, PREST, DEST, DST and DAST in the UK • These tests can identify problems before a child tries (and fails!) to learn to read. • They lead naturally to early support, on the ‘stitch in time’ principle. Early results are very promising! • These tests can identify both need and response to intervention, providing an effective and cost-effective solution