x

We are happy to announce that all trial registrations will now be issued DOIs (digital object identifiers). For more information, see here.
Empowering Staff to Enhance oral language in the early years
Last registered on September 23, 2019

Pre-Trial

Trial Information
General Information
Title
Empowering Staff to Enhance oral language in the early years
RCT ID
AEARCTR-0004379
Initial registration date
September 20, 2019
Last updated
September 23, 2019 4:16 PM EDT
Location(s)

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
Affiliation
Newcastle University
Other Primary Investigator(s)
PI Affiliation
University College London
PI Affiliation
Oxford University
Additional Trial Information
Status
On going
Start date
2019-05-01
End date
2020-12-31
Secondary IDs
Nuffield Foundation # EDO/43391
Abstract
Developing children’s oral language skills has been the focus of research and policy for a number of years. Social disadvantage is closely associated with language delay and language delay impacts on social emotional and behavioural development and the ability of children to access the curriculum. The extent of poor oral language skills in the population and the significant impact that this has on children’s long term outcomes points to the need for a systematic approach to the challenge at both universal and targeted levels of intervention.There are currently a number of oral language interventions in the public domain both within the UK and more widely, the majority of which are targetedselecting subgroups of children within schools and classes who have been identified as ‘at risk’. By contrast this trial evaluates a universal intervention (Talking Time http://www.stass.co.uk/publications/talking-time) focusing on children at most risk of experiencing poor language trajectories, comparing their performance in language development to those children in nursery classes experiencing business as usual. The participants in this study are attending nurseries in the most socially disadvantaged areas in London and the north east of England. All children in the study will be between three and four years years of age. The nurseries will then be randomised to a treatment and a business as usual arm. The treatment arm will consist of a manualised intervention, Talking Time, which targets a range of oral language skills and provides staff with tailored support about talking with children.

Unlike other interventions which focus on targeted language support, Talking Time was devised as a universal approach which aims to support language learning for all children in the classroom by embedding the activities within the day to day running of the settings and supporting staff to develop ways of talking with children. Effectively through manualised activities staff are scaffolded in using different materials and foci to support children’s language learning. This is particularly important in areas of social disadvantage where both the children’s oral language is less well developed and the input they are reported to receive in early years settings is more impoverished.
For educational settings a strength of Talking Time is that the activities are consistent with the early years curriculum and training of staff in effective ways to talk with children will, in theory, generalize to the setting more widely. Thereby benefitting the children who experience Talking Time and also increasing work force capacity. We explicitly test this hypothesis, unlike in other studies which focus on targeted interventions, by including setting measures and staff questionnaires.

The intervention has been found to be effective in a small trial in inner London and will be scaled up in this study and enhanced to include professional development to support adult-child oral language interactions within the activities and across the settings. The primary outcomes are changes in the children’s language skills and the language-learning environment of the classroom. Social emotional development and staff knowledge are examined as secondary outcomes. The effects of child and setting factors will also be included in the analysis to capture barriers to implementation.
External Link(s)
Registration Citation
Citation
Dockrell, Julie, James Law and Sandra Mathers. 2019. "Empowering Staff to Enhance oral language in the early years ." AEA RCT Registry. September 23. https://doi.org/10.1257/rct.4379-1.0.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details
Interventions
Intervention(s)
Intervention
Talking Time is an oral language intervention with, as a primary outcome, the language and communication development of preschool children. In this study it is delivered by appropriately trained educational professionals who have received specific professional development (PD) from the research team. There is also strong evidence to suggest that oral language skills impact on social, emotional development (SED) and so we examine this possibility by including a measure of SED as a secondary outcome.

The Talking Time arm of the study will be compared with a "Business as usual" arm. It is predicted that there will be no differential impacts on the non-targeted skills (non-verbal ability and phonology).

The intervention will run from January to April 2020.

Intervention Content: The manualised component of the intervention will be based on Talking Time. Three main activities will be included in the trial:
1. Acting Out involves a series of dramatic activities and games using target vocabulary.
2. Story Talk supports children in talking about the pictures in a book to draw inferences using language and to develop complex vocabulary and grammar.
3. The Hexagon Game provides children with a visual stimulus to support the construction of personal oral narratives.
Acting out and Story talk will be implemented over the full intervention period. The Hexagon game, which emphasises narrative language, will be added to the range of activities in week 8 of the intervention period (as per original protocol).

The PD element of the trial will include three components:
1. Materials and training to prepare practitioners to use the intervention activities and materials, and make the language-learning principles behind them explicit. TT already includes training guidance and videos of staff carrying out the activities and this will serve as the starting point, supplemented by additional DVD footage to be developed and piloted.
2. Support sessions from mentors to provide feedback on TT practice, with a focus on recognition and acknowledgement of specific language supporting behaviours, with the aim of increasing both teachers’ understanding of the principles behind these behaviours and the likelihood that they will apply them within their practice with other activities.
3. In addition to mentor-supported analysis of practice, teachers will also be encouraged to discuss and share practice with each other, to build a strong "community of practice" using the TT intervention.

Delivery: In the autumn term (2019) all classes will be provided with a whole school CPD session on the impact of early oral language delivered by the PIs. This session will also outline the assessment measures that will be used in the classes. The intervention group will receive a further three delivery sessions (one covering each of the three activities and associated principles) and will be provided with exemplar materials to support the activities. Training will be provided by independent appropriately qualified trainers.

Posttest data will be collected in the summer term (2020) on completion of intervention. This provides the necessary time gap to conform with test restrictions for the standardised measures.

End of intervention:
To acknowledge their involvement all participating schools will be given a book token following successful collection of the follow-up data. Following the post-test data collection BAU classes will be provided with the intervention materials and the option of a training session by both the PI or CoIs. A final interactive conference will be held for all participating schools in the north and the south.
Intervention Start Date
2020-01-01
Intervention End Date
2020-04-30
Primary Outcomes
Primary Outcomes (end points)
Child level: Oral language (vocabulary, narrative and receptive language) will be measured by a battery of standardised assessments that are commonly used in the literature and administered to children individually. These measures have been standardised on a representative UK population and the BAS3 provides Rasch ability scores to ensure that progress can be benchmarked in relation to item difficulty. Using the BAS3 for both verbal and nonverbal outcomes has the added benefit of consistency of standardisation across key targeted and non-targeted skills.
The battery consists of: Naming Vocabulary and Verbal Comprehension subtests from the British Ability Scales: 3rd edition (BAS3; Elliot & Smith, 2012); a bespoke list of vocabulary words taken from the Object and Action Naming Battery (Druks & Masterson, 2000), sentence repetition from the Grammar and Phonology Screening test (GAPS; Gardner, Froud, McClelland & van der Lely, 2006) and the Bus Story test (Renfrew, 1995).
Additionally, parent report of child language will be measured by a shortened version of the Language Use Inventory (LUI; O’Neill, 2009)..
Setting level: Quality of classroom language learning environment, interactions and opportunities measured by observation using the Communication Supporting Classroom Observation Tool (CSCOT) (Dockrell et al., 2012), and the language scales of the Early Childhood Environment Rating (ECERS) scales (ECERS-3; Harms et al., 2015), the ECERS-E (Sylva et al., 2011) and the Sustained Shared Thinking and Emotional Wellbeing (SSTEW) scale (Siraj et al., 2015). Staff knowledge and attitudes will be measured by a bespoke questionnaire


SECONDARY OUTCOMES
Setting measures: Following previous work (Mathers et al., in progress), a subset of relevant scales from the ECERS-3, ECERS-E and SSTEW rating scales will be complemented by the communication supporting classrooms tool (Dockrell et al., 2015) which examine language-learning opportunities and interactions.

Social emotional behavioural development: The Strengths and Difficulties Questionnaire (Goodman, 1997) is our proposed measure given its strong psychometric properties and availability in the range of languages. Parents and teachers will be asked to complete the questionnaire.

Primary Outcomes (explanation)
Children to establish whether the target children's language skills change to those of children in the comparison group receiving standard schooling or "business as usual".

None of the outcomes will be constructed de novo but the LUI identified above will be simplified somewhat to make it more relevant and will include an additional column to tap into performance in the home language in addition to English.
Secondary Outcomes
Secondary Outcomes (end points)
Parent reported social behaviour will be measured by the Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997).
Setting measures: Following previous work (Mathers et al., in progress), a subset of relevant scales from the ECERS-3, ECERS-E and SSTEW rating scales will be complemented by the communication supporting classrooms tool (Dockrell et al., 2015) which examine language-learning opportunities and interactions.
Secondary Outcomes (explanation)
The secondary measures relate to aspects of the classroom communication environment and the socio-emotional development of the children.
Experimental Design
Experimental Design
The overall aim of the study is to ascertain whether, in areas of disadvantage, training staff to use a manualised intervention in combination with PD about language development improves children’s oral language skills in comparison with business-as-usual (BAU); and whether it can impact on overall quality of the language-learning environment beyond the intervention sessions themselves.

The project started in May 2019, nurseries were identified in June/July 2019. Parental consent will be gained in September/October 2019. Data collection will run from September 2019 to November 2019. Initial training sessions will start in December 2019. The intervention and mentoring will run in the Spring term of 2020 (January to March). The children will be retested in the summer term of 2020. The data analysis, report writing and dissemination will take place between July and December 2020.
Experimental Design Details
Not available
Randomization Method
This study is a school-level cluster superiority randomised controlled trial. The randomisation process will follow ‘CONSORT guidelines for reporting of RCTs’ include cluster extension (see http://www.consortstatement.org/extensions?ContentWidgetId=554).

Nursery classes will be randomly allocated to the following conditions in approximately equal proportions (by location but not by setting because they vary in size):
• Intervention: Receipt of Talking Time training in Spring 2020
• Waitlist: "Business as usual" until after outcome data collection in Summer 2020

Consenting nurseries (clusters) will be randomly numbered and then those numbers (nurseries) will be randomly drawn and the nurseries then randomised to the intervention or control groups using the minimisation* technique, stratifying by; geographic area (North and South), and IDACI with a median split. These data will be collected at the initial assessment of the children prior to randomisation.

The randomisation will be carried out bunder the guidance of but not by the project statistican (Robert Rush) so that he remains truly independent of the allcoation when it comes to the analysis. The randomisation will be carried out based on data provided by the research team.

Allocation will be single blind with concealment at the assessment level. Research Assistants collecting the outcome measures at baseline and follow-up will be blinded to the allocation of nurseries. Those carrying on the training of the nursery staff will be aware of the randomsation status of the nurseries concerned.

We will carry out randomisation within stratification blocks to reduce the risk of imbalance on important characteristics between our resulting treatment and control groups. These stratification blocks will be formed by balancing approximately equally-sized groups in our two geographical areas.

We will carry out randomisation within stratification blocks to reduce the risk of imbalance on important characteristics between our resulting treatment and control groups. These stratification blocks will be formed by balancing equally-sized North and south groups.
Randomization Unit
This is a cluster randomised study with the "setting" (nursery) as the unit of randomisation.
Was the treatment clustered?
Yes
Experiment Characteristics
Sample size: planned number of clusters
Sample Size: Sample sizes are estimated from current DfE data about average setting size and previous research. Maintained classes have an average of 28 pupils in a nursery class. As this is a universal intervention that is part of routine practice previous work suggests that between 60 and 80% of the parents will opt in to the collection of child level data and continue for the duration of the project. Our target number of pupils per setting is 20 for collecting child level data. Language outcome data is not parent dependent but we anticipate that behaviour data will be less, as it depends on parents/teachers completing questionnaires (see section on quality assurance). Given previous similar studies offering universal interventions, we anticipate in addition an attrition rate of 10% of classes (see Bleses et al, 2017).
Sample size: planned number of observations
640 participants with 320 in each of the two locations, 36 classes, split equally between the two sites (London and the Teesside) with each child being assessed before (September to December 2019) and after (April to July 2020) the study.
Sample size (or number of clusters) by treatment arms
Two arms, 36 clusters.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
The initial trial of TT achieved effect sizes (Cohen’s d) between .15 to .68 for oral language outcomes, with 90% participation in the intervention setting. A power calculation using G*power to detect an effect size D>.20 including baseline measures of structural language, pragmatic language and behaviour with an alpha set at p<.05 suggests that the trial requires N = 600 participants. Sampling 20 three year olds per setting and anticipating a further 10% loss off classes over the period of the project our intended recruitment is 36 classes (of which 18 are the BAU comparison).
Supporting Documents and Materials
Documents
Document Name
Dockrell,Stuart and King 2010
Document Type
other
Document Description
This paper provides the original rationale and outcomes for the present trial.
File
Dockrell,Stuart and King 2010

MD5: 728b78e335d988e1ba3bc5fad52103f6

SHA1: d7f623f9ac39b3ae6dc9e616fdbd08f3a6c18f17

Uploaded At: June 26, 2019

IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
University College London
IRB Approval Date
2018-09-07
IRB Approval Number
Z6364106/2018/08/108 social research