Professor Julie Dockrell

 

The Project

Project Name:  Language-learning needs on school entry:  profiling phonology, morphology and narrative

Summary

Speech and language difficulties are the most common disability of early childhood, yet they are the least well detected. Significant numbers of children enter school with poor oral language skills. Yet teachers report a lack of reliable and valid tools to identify children oral language problems and to profile language-learning needs. A number of potential tools exist for use but these are limited because of time/resources involved, need for specialist qualifications or lack of psychometric data to support their use. The GAPS test is a notable exception and, importantly, met the criteria to be included in the Education Endowment Foundations database of recommended assessment tools.

  
For pupils under the age of five years regular monitoring of language is recommended, as one-off assessments have limited power to predict later performance. By the time children enter formal education there is more stability; where language at five years, both as reported by parents and as measured by cognitive assessment of vocabulary, are powerful predictors of teacher-identified language-learning needs at seven and 11. To date where studies report stability receptive vocabulary has been the primary focus. Recent research suggests that growth in morphology and narrative skills may be more important indicators of language benchmarks in the early school years.

 
The proposed project aims to examine the ways in which van der Lely’s GAPS test, in combination with a measure of narrative language, could meet teachers’ needs to identify language difficulties and to profile language performance through reception and into Year 1.
Currently there is robust evidence to justify the use of the GAPS for assessing phonology and morphology when used by skilled practitioners. Evidence about its predictive validity and relationship with school-based assessments is missing.

  
The current project aims:

  
1.     To examine:

 

(a)    growth curves in children’s performance on the GAPS and narrative language; sampling performance at three time points

(b)    the extent to which the GAPS and narrative concurrently and over time are differentially associated with educational measures (including baseline assessment, the phonics screen and school measures of language/literacy)

2.     To contribute to the current research examining stability of children’s language trajectories from the age of 5 by extending the focus to include morphology, phonology and narrative and the extent to which trajectories can predict language-learning difficulties.
An 18-month project is proposed examining performance of pupils through Reception and Year 1.  Power analyses indicate that a sample size of 200 pupils will allow growth curve modelling and provide sufficient power to examine relationships with Baseline measures, the phonics screen and school measures of language/literacy.  A sample of 240 pupils (to allow for attrition) stratified by Socioeconomic Status and balancing urban and rural locations, including children with EAL and those in the mainstream with special educational needs will be recruited. Both Baseline school measures and measures of language and literacy will be collected using standardised tests.

 

In sum, this project builds on van der Lely’s seminal work to inform theory and educational practice.
 

 

 

Professor Charles Hulme and Dr Gill West

 

The Project

Project Name: Standardisation of ATLAS and GAPS computerised oral language assessment tools

Summary 

Oral language skills are a critical foundation for literacy, educational success and psycho-social wellbeing.  Accurate, reliable and easy-to-use tests are needed to identify children with language learning difficulties.

The Grammar and Phonology Screening test (GAPS: Van der Lely, Gardner, McClelland, & Froud, 2007) is a short, child-friendly, paper-based test that assesses sentence and nonword repetition.

 
We have developed a short, but reliable, oral language assessment for children (The Automated Test of Language AbilitieS – ATLAS), which comprises 4 subtests (expressive vocabulary, receptive vocabulary, sentence repetition and listening comprehension).  The test has been developed as an App, running on a tablet, designed to be administered by teachers or teaching assistants.  The App can be administered in under 10 minutes and scoring is automated via a secure website.

Results from a preliminary sample of 305 children aged 3 to 6 years show that the ATLAS subtests are reliable, correlate highly with relevant subtests from the CELF battery (Wiig, Secord, & Semel, 2004), and exhibit high factor loadings on a single “language” factor.
The current application is for funding to develop a tablet-based App version of the existing GAPS test and to standardise both the ATLAS and the GAPS Apps in a nationally representative sample of 1000 children aged 3 to 7 years (250 children in each of 4 year groups from Nursery to Year 2).  Rasch scaling will be used to refine item selection and scoring algorithms developed to generate separate subtest scores and overall language scores (age-standardised scores and age-equivalent scores).  We will validate the App scores in a random sub-sample against well-standardized tests of language function (recalling sentences and expressive vocabulary subtests from CELF Preschool-2: Wiig, Secord, & Semel, 2004).
The GAPS and ATLAS tests both contain a sentence repetition subtest. The GAPS test includes a nonword repetition test that is not included in the ATLAS.  If this application for funding is successful, the work could be used in different ways.  Option 1 – we hand back the GAPS App and data to the remaining members of the development team to use as they see fit.  Option 2 – we combine the GAPS and ATLAS tests into a single test, giving appropriate acknowledgement to those parts of the new test derived from the GAPS (we would produce a single Sentence Repetition subtest using whichever of the items from the 2 existing scales had the best psychometric properties). 

 
The creation and publication of either one or two automated tablet-based Apps to assess language skills will have important implications for the identification of children with language learning difficulties.  The new test(s) will be an important research and clinical resource and play a vital role in allowing education professionals (teachers and teaching assistants) to identify children who require language support and remedial teaching.

The budget for this project is £140,000, to include all staffing costs, GAPS App development, tablet computers for screening and travel.

​

​​

Dr Fiona Kyle

 

The Project

Project Name: Using the GAPS to identify pre-school deaf children at risk of later literacy difficulties

Summary

Learning to read is a challenging experience for most deaf children, due to impoverished auditory access with consequences for spoken language acquisition.  As a result, many are significantly delayed in their reading ability.  Our previous research has shown that vocabulary and phonological skills are strong predictors of reading outcomes in school-age deaf children. Because of the typical severity of later reading delays, it is crucial to identity which deaf children are likely to have reading difficulties BEFORE they start learning to read.  Early identification will help ensure they have appropriate support, before beginning to read.

The aim of the proposed study is to determine whether we can predict which deaf children will have later literacy difficulties from their pre-school language and phonological skills.  We know that pre-school emergent literacy skills (e.g. letter-sound knowledge, phonological awareness, print knowledge and vocabulary) are strong predictors of later reading difficulties in hearing children.  In comparison, little is known about the predictive role of these early skills in deaf children.  The proposed study will investigate the validity of using the Grammar and Phonology Screening test (GAPS) along with tests of emergent literacy skills to identify strengths and weaknesses in the pre-reading skills of pre-school deaf children, in comparison with same-age hearing children.  In addition, we will explore the predictive validity of assessing pre-school deaf and hearing children using the GAPS to identify later reading outcomes and difficulties.  The GAPS is a ten-minute grammatical and phonological screening test that has been shown to reliably predict later reading and language difficulties in hearing children but has not been previously used with deaf children.  If the GAPS is valid for use with deaf children, it will reveal more about the role of early grammatical difficulties in later literacy and it could be utilised as a time-efficient screening toolkit for practitioners working with deaf children.

We will conduct a longitudinal study which follows a group of pre-school deaf and hearing children over two years. We will assess 40-50 pre-school deaf children (4 year-olds) who communicate through spoken language and 50 pre-school hearing children and re-assess them in reception and year 1.  At the beginning of the study, children will complete brief and child-friendly assessments of emergent literacy skills and the GAPS.  In reception and year 1, they will repeat these tasks, along with assessments of reading and spelling.

We will partner with the Elizabeth Foundation (http://www.elizabeth-foundation.org), a national UK charity providing on-site and outreach support and pre-school education for large numbers of deaf children and their families.  We will recruit families through the Elizabeth Foundation and through our established contacts with pre-school education services for deaf children.  

The Elizabeth Foundation will also be an impact partner in disseminating study findings.  The study will generate two academic publications in peer-reviewed journals.  The team has research expertise in reading and language development in deaf children and conducting longitudinal studies.  

 

 

Professor Cristina McKean

 

The Project

Project Name: Testing the efficacy of Building Early Sentences Therapy and uncovering mechanisms for change: a theoretically informed language intervention for pre-school children with language disorder

Summary

This study aligns with Professor van der Lely’s aims for her legacy “to further the quest for solutions about the nature of interventions that best succeed in helping children with developmental language impairments.”

We propose to conduct a Cluster Randomised Controlled Trial (RCT) of Building Early Sentences Therapy (BEST) http://www.buildingearlysentencestherapy.co.uk: a language intervention for children aged 3;06-5;06 with severe language difficulties informed by ‘usage-based’ linguistic theory. We will compare BEST to one of the most widely used interventions: the Derbyshire Language Scheme (DLS). DLS has no underpinning theory but has many similarities to BEST in terms of delivery, target language structures, and age of the children. This allows for comparison to an intervention delivered with the same dosage, delivery context, level of treatment fidelity and similar resources. We will conduct this trial in English and also complete a series of single case studies applying BEST principles to languages other than English (LOTE), providing therapy to children with language disorders in their home language.

We aim to determine whether

1) BEST is effective;

2) the underpinning principles of BEST can be applied successfully across languages;

3) it is the specific learning mechanisms exploited by BEST, rather than ‘therapy-general effects’ or language-specific effects which promote change

BEST aims to improve children’s use and understanding of two, three and four element sentences (e.g. The girl is jumping; the boy is eating a banana; the baby is putting the cup on the table).  Rather than encouraging rote-learning of language structures BEST aims to promote abstract and flexible representations of sentences: an accomplishment thought to accelerate future language learning. The underlying principle is that the nature, distribution and quantity of the language input a child hears is central to the process of acquisition. By systematically manipulating the input, BEST aims to promote children’s use of learning mechanisms including pattern finding, analogy, and categorisation, thought to support language acquisition.

Methods
Cluster RCT in English:

12 schools will be recruited and randomised to BEST or DLS.  Research Associates (RAs) will deliver interventions in partnership with schools for 3 terms (Total N=~144 children aged 3;06-5;00 with severe language difficulties).  RAs ‘blind’ to treatment arm will complete language and functional communication assessments pre and post-intervention and at 4 months follow-up (New RDLS, BEST assessment, DLS assessment, GAPs, FOCUS).

Case-series in LOTE:

Partner schools with high numbers of children who speak LOTE will be identified and 2 target languages chosen.  Assessment and therapy materials will be developed drawing on native speaker consultations and team expertise.  Nine children aged 3;06-5;00 with severe language difficulties in their home language will receive BEST delivered in partnership with parents or bilingual co-workers in schools. Recommendations for best practice in single case methods will be applied and children’s rate of progress before, during and after intervention compared.

​

Dr Helen Spicer-Cain

 

The Project

Project Name: SibSpeak Follow-Up Study

Summary 

Developmental Language Disorder (DLD) is a relatively common communication impairment, affecting around 7.6% of children at school starting age (Norbury et al, 2016).  Evidence suggests that most parents of children with DLD identify concerns about their children’s communication skills before the age of two (Rannard et al, 2004), and yet many children do not access speech and language therapy services until they are of school age (e.g. Rannard et al, 2004; Broomfield & Dodd, 2011).  This delay in identification is particularly concerning in light of evidence that many children with DLD experience psychosocial, emotional, behavioural and educational difficulties (e.g. Snowling et al, 2001; Conti-Ramsden & Botting, 2008; Law et al, 2009; Whitehouse et al, 2009) and that language intervention before the age of five can be successful in improving children’s communication skills (e.g. Roberts & Kaiser, 2011; Law, 2017).  There is therefore a need for research which addresses the question of how children with DLD can be identified earlier.

Assessment data, including standardised language assessment, phonological measures, parent questionnaires and novel dynamic assessment measures (receptive language, motor imitation, point following, turn taking and requesting), were previously taken from 92 children at 8-22 months of age.  This included 17 children at high risk of social communication disorder (later-born siblings of a child with ASD or other social communication difficulty) and 22 children at high risk of language disorder (either later born siblings of children with language difficulties or children of parents with diagnosed dyslexia), in addition to 53 control children whose first-degree relatives had no history of language or literacy difficulties. A sub-sample of the children were followed up at the age of 2-3, and dynamic assessment measures taken in infancy were found to be significant predictors of their language ability as pre-schoolers.  These measures have the advantage of being easy to administer in children’s home settings without significant training.  They therefore show promise as a means of early identification of children who will go on to have language difficulties.

The proposed project would seek to follow up this sample in the early school years and evaluate the ability of measures taken in infancy to predict language ability at age 5-6, with a particular focus on measures that predict later DLD. 

 

This would require around £90,000 to cover salary costs, purchase of assessment materials, travel for data collection and dissemination of results via conferences. 

 

Children would be seen for two sessions and tested using measures of receptive and expressive language and school readiness, with a focus on grammar, phonological awareness and performance on marker tasks known to be sensitive to DLD (sentence repetition and non-word repetition).  Information would also be sought from parents on speech and language therapy involvement and parent questionnaires would be used to identify co-morbid social, emotional or behavioural difficulties.  It is hoped that this research will contribute to a growing evidence base on early identification of DLD, which will allow more children to access support at a younger age.

 

​​​​​

​

​PhD research projects​

​

 

Keira Radice - Birmingham City University 

Continuing studies​​​

​

International research has shown Speech, Language and Communication Needs (SLCN) are disproportionately high within the youth justice population relative to community samples (Anderson et al., 2015; Chow et al., 2022; Royal College of Speech and Language Therapy (RCSLT), 2017). One SLCN, Developmental Language Disorder (DLD), has been of particular interest in recent literature (Anderson et al., 2015; Winstanley et al., 2021).  

 

DLD is a diagnosis given to a person who has difficulty talking (expressive language) and/or understanding language (receptive language) (Bishop et al., 2017). Young people (YP) accessing Youth Justice Services (YJS) will be exposed to a range of experiences which rely heavily on expressive and receptive language skills. This presents many barriers for YP, as they may not completely understand what is happening, fully know what is expected of them in the YJS or know how to successfully participate in rehabilitation or conditions set by their imposed court orders (RCSLT, 2017).  

 

Regardless of this emerging research, Speech and Language Therapy (SLT) interventions that target the aspects of language that could support the engagement and participation of YP within YJS are not readily available or funded. To become available, there is a need to prove the efficacy and benefit of SLT interventions at a policy level, so there is a subsequent desire to invest in SLT services.   

 

This research therefore seeks to examine the effectiveness of a language intervention, using an evidence-based framework, for YP with DLD who are accessing a local YJS Team.    

 

 

Caitlin Holme - UWE Bristol 

Caitlin has now been awarded her PhD - summer 2024. Congratulations!

 

A day in the life of 12 toddlers: studying children’s everyday interactions to inform parent-

child interaction therapy

Authors: Caitlin Holme1, Yvonne Wren2, Sam Harding3, Sue Roulstone4

Affiliations:

1. University of Bristol/Bristol Speech and Language Therapy Research Unit (BSLTRU)

2. University of Bristol/BSLTRU/Cardiff Metropolitan University

3 / 4. BSLTRU

Keywords: parent-child interaction, DLD, naturalistic

 

Abstract:

 

Introduction

Parent-child interaction therapy (PCIT) is a common intervention used with children at risk of

developmental language disorder. This approach involves working with parents to modify

their interactions in ways thought to be conducive to language development (Falkus et al.,

2016). Qualitative research exploring parents’ views of PCIT reveals that it is most successful

when strategies are embedded in families’ everyday routines (O’Toole et al., 2021). This

paper will explore preschool children’s everyday interactions in naturalistic settings, with the

aim of informing the theory behind PCIT.

 

Method

Participants were twelve families from diverse backgrounds with a child aged 2 ½ to 4 years

old. On a day of their choice, parents took photographs of their child’s activities and the

child wore an audio device which recorded their interactions. The parent then participated

in a qualitative interview to discuss their activities and interactions, incorporating the

photographs and automated recording analysis.

 

Results

Automated timelines showing moments of high and low interaction were annotated with

information about families’ activities. Photographs and quotations from the interviews

showed variation in how families engaged

in activities, and the times of day that provided opportunities for interaction.

 

Conclusions

We will explore variation in the activities that families did and the opportunities for

interaction that arose. Clinical implications for how we conduct PCIT will be considered,

including the importance of understanding the socio-cultural context of parent-child

interaction.

 

To access the Project Research Page please click here.