Early Intervention for Children with Language Difficulties An

Early Intervention for Children with Language Difficulties: An Evaluation of Two School Based Intervention Programmes Claudine Crane, Margaret J. Snowling, Julia Carroll, Fiona Duff, Elizabeth Fieldsend, Jeremy Miles & Charles Hulme Universities of York and Warwick, UK

Outline Theoretical rationale n Design of study and overview of Programmes n Participant selection n Training of Teaching Assistants n Findings n Implications for theory and practice n

Background n n n According to the simple model, Reading Comprehension depends on both decoding and listening comprehension Muter et al (2004); different language subskills underpin different component reading skills n Phonological Awareness + Letters -> Decoding; n Vocabulary + Grammar -> Read Comp Children with speech-language impairments at high-risk of RD (e. g. Catts et al. , 2005; Snowling et al. , 2000)

Intervention n Evidence that early intervention programmes that train phonemes and letters in context of reading can facilitate reading development (decoding) in at-risk children (Hatcher et al 2004; Hindson et al. , 2005); Less evidence regarding the role of vocabulary and grammatical instruction Question addressed by this study: n n is it possible to improve the development of vocabulary and grammar skills in at-risk children? how do such training programmes differ in their effects from phonological training programmes?

Design n n n Evaluation of two interventions designed for children with speech and language difficulties in mainstream schools to be delivered by trained teaching assistants (TAs) Oral Language Programme and Reading with Phonology Programme Randomised Controlled Trial (following the Consort guidelines) 20 -week programme 4 test phases: pre-test, mid-test, post-test and maintenance test Investigators blind to group membership

December 2004: N = 960 children in 23 schools screened for participation in early intervention project. 3 schools excluded (n = 100): performance too high January 2005: n = 200 children selected from remaining 20 schools (n = 861) for further testing n = 4 children unavailable for testing Allocation Feb 2005: n = 160 children selected to take part in the intervention programmes (n = 8 from each school) and randomly allocated to one arm of the intervention project. 1 school withdrawn from programme Total n = 45 Selected n = 8 (4 children from each arm) n = 17 children replaced following discussion with teacher Mar 2005: 152 children seen for pre-testing. Intervention programmes begin. Reading with Phonology N = 76 allocated to intervention N = 75 received allocated intervention N = 1 did not receive allocated intervention – moved schools but maintained follow-up Reading with Phonology Discontinued intervention N = 8 – moved schools Lost to follow-up N = 3 Analysed n = 71 Excluded from analysis n = 2 Incomplete dataset Oral Language N = 76 allocated to intervention N = 76 received allocated intervention Follow-Up Analysis Oral Language Discontinued Intervention N = 1 – moved schools Lost to follow up N = 1 Analysed n = 75

Participant Recruitment and Attrition December 2004: N = 960 children in 23 schools screened for participation in early intervention project. 3 schools excluded (n = 100): performance too high January 2005: n = 200 children selected from remaining 20 schools (n = 860) for further testing

January 2005: n = 200 children selected from remaining 20 schools (n = 860) for further testing n = 4 children unavailable for testing Allocation Feb 2005: n = 160 children selected to take part in the intervention programmes (n = 8 from each school) and randomly allocated to one arm of the intervention project. Mar 2005: 152 children seen for pre-testing. Intervention programmes begin. 1 school withdrawn from programme: Total n = 45 Selected n = 8 (4 children from each arm) n = 17 children replaced following discussion with teacher

Mar 2005: 152 children seen for pre-testing. Intervention programmes begin. Reading with Phonology N = 76 allocated to intervention N = 75 received allocated intervention N = 1 did not receive allocated intervention – moved schools but maintained follow-up Reading with Phonology Discontinued intervention N = 8 – moved schools Lost to follow-up N = 3 Analysed n = 71 Excluded from analysis n = 2 - Incomplete dataset Oral Language N = 76 allocated to intervention N = 76 received allocated intervention Follow-Up Analysis Oral Language Discontinued Intervention N = 1 – moved schools Lost to follow up N = 1 Analysed n = 75

Participants (N=146) Measure Mean Stdev Age (mths) 58. 47 3. 33 Picture Naming (std sc) 6. 16 1. 37 Vocabulary (std sc) 6. 42 2. 20 Word Reasoning (std sc) 7. 62 2. 33 Block Design (std sc) 6. 86 3. 13 SDQ Total Deviance: Normal 61% Borderline 16. 4% Abnormal 22. 6% SES (Free Sch Meals: n=130) 24%

Teaching Assistants Teaching assistants selected by schools n Attended 4 day intensive training programme n 2 Refresher days n Fortnightly tutorials n On-site tutorials n

Structure of the Programmes n n Programmes conducted over 2 x 10 week periods Following initial introduction week, teaching was broken into 3 week blocks consisting of two teaching weeks and one consolidation week Each week consisted of alternating daily group sessions or individual sessions Repetitive session structure – familiar routine, positive reinforcement

Programmes n Oral Language - Vocabulary development - Speaking - Listening - Narrative production - Comprehension - Question generation n Reading with Phonology - Training in letter sound knowledge (Jolly Phonics) - Oral phonological awareness - Reading books at easy and instructional levels - Sight word vocabulary development - Letter formation

Programmes n Oral Language - Vocabulary development - Speaking - Listening - Narrative production - Comprehension - Question generation n Reading with Phonology - Training in letter sound knowledge (Jolly Phonics) - Oral phonological awareness - Reading books at easy and instructional levels - Sight word vocabulary development - Letter formation

Measures Language Skills § § § § Reading Comprehension Listening Comprehension Specific Vocabulary Action Picture Test Bus Story WISC III Picture Arrangement Information Carrying Reading and Phonological Skills n n n n Early Word Reading Letter Knowledge Spelling Reading Accuracy Segmenting and Blending Sound Isolation Articulation

Measures Language Skills § § § § Reading Comprehension Listening Comprehension Specific Vocabulary Action Picture Test Bus Story WISC III Picture Arrangement Information Carrying Reading and Phonological Skills n n n Early Word Reading Letter Knowledge Spelling Reading Accuracy Segmenting and Blending Sound Isolation

Mode of Analysis n n n Data are clustered: 4 children per arm; two arms delivered by each TA Complex samples analyses giving robust estimates and CIs (SPSS 14) Primary outcomes n n n n Vocabulary Grammar Phoneme Awareness Letter Knowledge Word Recognition Reading Comprehension Covariates : age, gender, autoregressor (when available)

Relative Advantage of Language Gp in zscore units (95% CIs)

Relative Advantage of Reading with Phonology Grp in z-score units (95% CIs)

Summary n Both intervention programmes were effective in promoting basic skills that underlie reading comprehension n n Vocabulary and grammatical skills fostered better by language program (effect sizes. 25 -1. 02) Word-level reading skills and phoneme awareness fostered better by early literacy program (effect sizes. 21 -. 71) Biggest effects of training on receptive vocabulary (1. 02) and segmentation/blending (. 71) Neither program had significant effect on reading comprehension (effect size =. 19) at this early stage in development

Predictors of outcome n Explored two further predictors of children’s outcome n n Behaviour (SDQ total deviance score) Socio-economic circumstances • Post code index of dis/advantage • Free school meals n n Controlling for behaviour had no effect on findings Significant influence of ses on outcome

Effect of SES? When SES controlled training effects remained significant for early literacy and phonological measures, and vocabulary n No longer significant effect of training on grammar (APT; Bus story)nor sequencing – picture arrangement n

Conclusions n n n Focused intervention programs can be delivered successfully by teaching assistants to 5 and 6 -year-old ‘at risk’ children Such programmes can foster the basic skills that underpin word-level and text level reading skills Programmes emphasizing early literacy versus oral language skills have differential effects

Conclusions 2 In terms of effect size, specific vocabulary and phoneme awareness skills appear to be the most ‘trainable’ skills n Preliminary evidence that social class moderates gains in grammar but not in vocabulary n

Thank You n n Funders: Nuffield Foundation, North Yorks County council Sponsors: Jolly Phonics; Black Sheep Schools: Pupils, TAs and teachers Assistants: Naomi Meredith, Nicky Vowles, Rachel Harlow, Debbie Gooch, Ros Francis, Dimitra Ionnau, Lisa Henderson, Lizzie Bowen, Natalie Falkinder, Sarah Edwards, Emma Truelove, Kim Manderson, Jodie Unau, Michelle Cargan, Pam Baylis, Rachael Mc. Cool, Elisa Romeo, Meesha Warmington, Poppy Nash, Janet Hatcher