Unfavourable opinion from NHS ethics. Due to low recruitment from NHS sources in another study, it was decided not to reapply but to withdraw the study.

The promotion of language and communicative development in the early years is extremely important. Children who enter school with good language skills have better educational and economic success. This study is part of a large project across Liverpool, Manchester and Sheffield Universities to determine how shared reading promotes child language development, and use this knowledge to make it an effective language boosting tool for children across the whole socio-economic spectrum. The overall project includes:

• observational studies to identify what language boosting behaviours are responsible for shared reading's effectiveness, and how parents from different socio-economic groups use these behaviours during shared reading;
• intervention studies to evaluate packages designed to train parents in the use of specific language boosting behaviours during reading;
• a qualitative exploration of the reasons people may not read with their children.

This study will provide training to parents on how to develop their children's attention to the features of words while reading books with them. The research questions are:

i) Is specific training focused on the sound properties of words during shared reading more effective at developing children's phonological awareness and language than general advice on the importance of reading with children? ii) Do children with speech sound disorder and typically developing children respond differently to intervention? iii) To what extent are differences in training implementation and effects explained by socio-economic status?

Our participants will be parents and their children, aged 30-54 months, with a diagnosis of speech sound disorder. They will be recruited via speech and language therapy services in the North West. Data collection will be carried out by the research team in participants' homes, taking 3-4 hours in total over 2-3 appointments. The sessions will be audio-recorded; parents will complete questionnaires, and children's language and speech will be assessed with standardised and in-house tests.

GLOSSARY OF ABBREVIATIONS PLS-5 Preschool Language Scale: Fifth Edition UK PIPA Preschool and Primary Inventory of Phonological Awareness SES Socio-economic status 2;6 Age in years;months, i.e. two years six months TD Typically developing SLT Speech and Language Therapy REC Research Ethics Committee UREC University Research Ethics Committee

Participants Approximately 60 parent/child dyads will be recruited, where the child has a speech sound disorder. Legal guardians other than parents will be eligible; for simplicity we use the term 'parents'.

Expressions of interest from potential participants will be gained by the study team and consent procedures will be administered by the study team only. Participants will be recruited in the North West of England and in North Wales. In order to recruit enough participants of low SES, areas of disadvantage will be more heavily targeted for the first three strategies listed.

Parents will be made aware of the study, and asked to indicate expressions of interest, via the following routes:

• SLT services who will be approached to identify children and families from their current and recently discharged caseloads who may be suitable for inclusion in the project*
• advertisements in the media, with the help of University Press Officers, and distribution of leaflets and posters (see attached) in the community (nurseries, playgroups, mother and toddler groups, libraries and sports centres that host activities for young children, baby sign language groups, General Practitioner (GP) surgeries, pre-school libraries, health visitors, baby clinics, Children's Centres and similar). All posters and flyers will include the same text, including contact details for the research team (work details only).
• direct liaison with local community groups and professionals interacting with parents and children in the community
• personal visits to local community centres and groups
• word of mouth recruitment of participants' friends/families
• recruitment of volunteers who have directly contacted the University of Manchester to express an interest in this or other research.
• internet-based recruitment (including social media) with online sign-up
• parents who have already volunteered to participate at collaborators' and colleagues' laboratories, and have indicated their willingness to be contacted about other studies.

Following expressions of interest:

• the information sheet and sample consent form will be sent through the post or via email to potential participants

• at least 48 hours later, follow-up phone calls will be made to answer queries and check ongoing interest. During this phone call the researcher will discuss in detail with the family what the study involves, including their right to withdraw from the study, and reassure parents on procedures for data collection and storage that ensure confidentiality and anonymity. If the participant is happy to be involved, the researcher will run through an eligibility checklist to ensure the family are suitable for the study based on the exclusion criteria

• the researcher will recap this information the first testing session and the parent will, if willing, sign the consent form

• during this initial testing session, the family's suitability for the study will be checked again from the Family Questionnaire, and speech production screen for children with SSD.

  • Procedure for recruitment via SLT services SLT services will be approached to identify children and families potentially suitable for inclusion in the project. They will be provided with information sheets (standard and child-friendly) and consent sheets to give to families who are suitable. Families will be invited to read the materials and indicate willingness to hear more about/take part in the project by either a) responding directly to the researchers via contact details on the form or b) responding via their local SLT contact who will pass on contact details to the research team.

On contact, the research team will answer by telephone or email any queries raised by parents, and will run through an eligibility checklist to ensure the family are suitable for the study based on appropriate items from the inclusion/exclusion criteria set out above. If parents agree, researchers will arrange a visit to the home.

Consent Consent will be sought at the beginning of the first assessment session. If signed consent is obtained, data collection will commence.

All the adults taking part in this study (parents/guardians) will be able to give informed consent on their own behalf (see exclusion criteria above). The children taking part in this study will be aged between 3;6 and 4;6 years of age and are thus not deemed able to give informed consent; their parents or legal guardians will give consent on their behalf. Children will be asked to give verbal assent. Indications of distress will also be taken as non-consent during assessment; in such cases tasks will be re-commenced or re-scheduled or the child withdrawn from the study, according to parental preference.

Withdrawal If the assessments show the children/ parents do not meet the inclusion criteria they will be withdrawn from the study and their data not used. Families themselves will be told and reminded that they are free to withdraw from the study at any point without giving a reason. Families can withdraw in writing, in person, by email or by phoning any member of the research study team.

STUDY METHODS

Randomisation and blinding Randomisation will be determined using a computerised random number generator with equal randomisation to intervention and attention control, and a block randomisation design to ensure equal numbers of participants in each condition in rolling recruitment. The allocation sequence will be held by an administrative member of staff and the two researchers will call that person to obtain the group allocation when they have enrolled a participant in the study.

Participants will be blind to the arm of the study they are in. Both groups will receive a training session of similar length, one focusing specifically on sounds in words and the other offering general advice on reading with children. The researcher enrolling and training the participants will obviously know which group the participant is in. However, the post-intervention assessment will be carried out by a different researcher who is blinded to group allocation.

Assessments and training Assessments and training will take place by one of the named researchers, either in the family home or the University according to room availability and family preference. The assessments and questionnaires to be completed are listed below (not all of them require the child's involvement). It may be possible to complete everything on one long visit (estimated two hours). However , more likely, at parent request or with their agreement, assessments and training will take place over two appointments.

PLS-5: A standardized measure of language development. The investigators will use both the Auditory Comprehension and Expressive Communication scales to confirm language ability. Completion time: 20-30 minutes.

PIPA: a standardised assessment of phonological awareness. Completion time 15-20 minutes.

The Family Questionnaire: records relevant demographic information (e.g. socio-economic status, parents' education, family language use, ethnicity), devised for the ESRC-funded UK-CDI study. This allows the investigators to create a composite socio-economic variable in order to identify high and low SES families. Completion time: 10 minutes.

Home Life Questionnaire: a parent report questionnaire including frequency of storybook reading, library visits, number of children's books in the home, parental reading habits etc. Completion time: 5 minutes.

Children's Title and Author Checklists: These checklists indirectly measure a child's storybook exposure by assessing the parent's knowledge of children's book titles and authors (Children's Author Checklist). They each contain a mixture of real authors/title and foils. Completion time: 10 minutes.

Speech production screen: a 20 item picture-naming task to confirm presence/absence of speech sound disorder. Completion time: 5 minutes.

Syllable segmentation screen: a custom-designed assessment to assess the ability of a child to segment words into constituent syllables. Completion time: 5 minutes.

Training in reading strategies or habits: an interactive 45-minute session in which parents are trained in either

1. how to draw children's attention to the sound structure of words during the shared-reading of books which emphasise word form, and generalizing this into everyday conversation; or
2. the importance of reading with children, how language develops in children during the 3;6-4;6 period, and strategies for fitting reading into the daily routine.

Duration of involvement in the study Seven weeks, including assessment sessions and intervention phase.

ADVERSE AND SERIOUS ADVERSE EVENTS Adverse events are not considered to be above the level of everyday risk. Participants are not being recruited on the basis of any medical condition and the study does not involve any procedure that might cause an untoward or unexpected adverse event. All the methodologies used are established, standard tasks in the field of language acquisition research.

Recording and reporting adverse (AE) and serious adverse events (SAE) All adverse events will be recorded. Participants will be provided with contact details of the Chief Investigator and the Research Governance Office at the University of Manchester.

All non-serious AEs will be reported to the Chief Investigator (CI), Dr Anne Hesketh, and will be recorded by the study team. NHS Research Ethics Committee will be notified of AEs online via the IRAS system.

In the event of an SAE, details will be recorded and emailed to the CI within 24 hours, who in turn will immediately report them to the University's Research Governance Officer and the NHS Research Ethics Committee.

STATISTICS AND DATA ANALYSIS The statistics and data analysis for the overall project were subject to internal review at the University of Liverpool and scrutinised in detail by the ESRC reviewers.

Participant numbers are based on power calculations for the intervention studies of the over-arching project which are based on the following assumptions: Power = .80; Alpha = .05; Medium effect size: d = 0.5, f2= .15. The intervention studies are powered to check for an effect of intervention compared to control and to check whether, for participants in the intervention condition, socio-economic status (SES) explains intervention effect. To test for this, 55 participants per group are needed (to compare mean difference scores for intervention and control conditions with an independent t-test, and to run a regression model with difference scores as the outcome measure and SES as a continuous predictor).

A number of statistical tests will be carried out depending on the nature of the hypotheses. ANOVAs, regression models, mixed effects models, and correlations will be used. Statistics will be performed in Excel, SPSS and R.

REGULATORY ISSUES Ethics approval The study is being submitted to the NHS REC (families with children with SSD) for ethics approval. Ethics approval has already been gained from the UREC for families/children with typical speech development. It will be conducted in accordance with the recommendations for research on human subjects adopted by the 18th World Medical Assembly, Helsinki 1964 and later revisions.

Confidentiality The overall grant CI (Liverpool) will oversee the confidentiality of participants taking part in the study and is registered under the Data Protection Act. A data management plan is in place which conforms to the requirements of the ESRC and all participating Universities; it is the responsibility of the CI at each participating University to adhere to the Data Management Plan for their studies. All data will be pseudo-anonymised. Each participant will be given a participant identification code and all data (case report form, test scoresheets, video/audio files, questionnaires transcripts) will be identified using this code. A separate, password-protected datafile will allow members of the immediate study team only (but no others) to match participant code with participant identification information. Video data will be transferred from the camera to secure University servers on return to base. Data recorded on encrypted laptops will transferred to secure University servers on return to base. Paper copies of questionnaires will be labelled by code and kept securely according to ESRC and University data retention policies.

Insurance The University of Manchester will arrange insurance for research involving human subjects that provides cover for legal liabilities arising from its actions of those or its staff or supervised students, subject to policy terms and conditions.

Audits and inspections The study may be subject to inspection and audit by the University of Manchester under their remit as sponsor and other regulatory bodies to ensure adherence to Good Clinical Practice and the NHS Research Governance Framework for Health and Social Care (2nd edition).

END OF STUDY The study will end on 31/03/2018 when the funding comes to an end.

ARCHIVING

Data will be shared with the UK Data Archive in accordance with the funder's (ESRC) data sharing guidelines http://www.data-archive.ac.uk/. Data will be managed on three tiers:

1. Data that will be made fully publically accessible. This will include new anonymised corpora of transcripts that will be donated to CHILDES (http://talkbank.org/share/ethics.html) and the UK Data Archive.
2. Data that will be made publically accessible in fully anonymised summary form via the UK Data Archive (e.g. experimental datasheets in Excel or SPSS that allow others to replicate the analyses in published papers)
3. Data that will only be available to the immediate research team (e.g. raw uncoded data, video & audio recordings that do not have permission for data sharing, consent forms, any other data that allows the identification of participants).

DISSEMINATION POLICY

Results from this study contribute to the overall project dissemination plan:

Outputs for academic users in high impact interdisciplinary and specialist journals and via academic conference presentations.

Outputs for non-academic users to include newsletters for parents, reports for early years practitioners and policy makers, press releases, and invited conference talks. All these will be produced in collaboration with the ESRC LuCiD Centre (Manchester, Liverpool and Lancaster Universities) and will follow the LuCiD Centre's Communications Agenda. The Policy Implementation Strategy will communicate best practice and crucial findings to the local and central government bodies responsible for early years and health policy; The LuCiD for Professionals programme will establish a comprehensive communication strategy for the dissemination of findings to healthcare and education professionals; the Discovering BabyTalk programme will establish a comprehensive public outreach agenda for families and the wider public, to provide them with the best evidence-based advice, and the tools they need to promote language growth; the media strategy, co-produced with University press offices, will disseminate the current research to the general, scientific and trade press for health and education professionals.

Inclusion Criteria:

• known to speech and language therapists with a diagnosis of speech sound disorder OR recognised by parents as having less mature speech production than their peers AND perform below criterion on a brief speech sound production screening assessment

Exclusion Criteria:

• a known neurological diagnosis (such as Down Syndrome, Autism, Cerebral Palsy)
• born before 37 weeks gestation (premature)
• weighed less than 5lb 9oz at birth (low birth weight)
• a permanent impairment of vision (unless remediated by visual aids) or hearing
• parents have a learning disability which puts their children at risk of language delay and excludes the parents from giving informed consent on their own and on their children's behalf.
• exposure to another language (not English) for 1 day or more in a typical week (please note that this also excludes children of parents who do not speak English)