Background

The role that children's language plays in promoting their educational and social success is increasingly recognised. This is the case for children who have particular difficulties in acquiring speech, language and communication and in a more general sense for the entire population. The longer term impact of speech, language and communication impairments on a child's literacy, education, behaviour, employability and social inclusion has been reported in a number of clinical studies (Baker & Cantwell 1987; Catts, 1993; Clegg, Mawhood, Howlin & Rutter 2005; Cohen, Vallance, Barwick, et al. 2000; Johnson, Beitchman, Young et al. 1999; Rutter, Mawhood, & Howlin 1992; Stothard, Snowling, Bishop et al. 1998). Findings that show high proportions of young offenders with poor speech and language have also increased awareness of the links between language skills and educational and social success. (Bryan, 2004; Bryan et al, 2007). Concern has been expressed that children are entering school 'without the extended vocabulary and communication abilities they need for learning and for making friends' (Children's Minister Beverley Hughes, 2008) and head teachers surveyed by the Basic Skills Agency were of the view that children's language skills have declined over the previous five years (Basic Skills Agency, 2002). 'Every Child a Talker' has been set up to address these concerns and enhance children's language during the Early Years.

Despite growing reports of the negative outcomes associated with limited language skills, our knowledge about which particular contexts and factors give rise to impairments and limited language in the individual is still largely speculative. Children who present as late talkers show varying rates of resolution by school age, depending on the clinical samples recruited; awide range of risk and predictive factors have been identified including speech and language features, characteristics of play, gesture and social skills, the presence of otitis media, family history as well as socio-economic factors and parenting style(Olswang, Rodrigues, and Timler, 1998). However, how such factors play out in the individual is still unclear. In particular, the role of children's social environment is controversial. So for example, Wells (1985) studied naturalistic language samples of preschool children in Bristol and concluded that there was no evidence to support an association between parental social status and the language environment of the child. Conversely, Hart & Risley (1997) compared the naturalistic vocabulary environment of children up to the age of 3 years old, in US families of differing social status and showed a threefold difference between families from professional backgrounds with those on welfare; these differences were still apparent when the children were aged 9 and 10 years old. Furthermore, it is not clear if there are factors which act as moderators that facilitate language acquisition in spite of adverse environmental conditions and in what conditions these might operate or indeed if language itself is a moderator of social adversity. In a study of educational resilience, Schoon et al (2004) noted that protective factors seemed to operate more with socially advantaged families. They concluded that protective factors are context specific.

Luthar et al (2000) note a number of difficulties with the investigation of resilience, in particular, definitional problems, the heterogeneity in how risk and resilience play out in the individual and the instability of resilience. Within the field of language, there have certainly been definitional issues: language impairment itself is a variable construct with the interface with normal distribution being controversial. Further, Schoon et al (2004) criticize the use of single indicators such as parental employment to identify levels of social risk. The Avon Longitudinal Study of Parents and Children (ALSPAC) provides a unique opportunity to investigate the role that language plays in the education and social success of individual children. The cohort has a wealth of data about the individual children's psychological and linguistic development, their family and social context as well as their educational outcomes both in terms of specific measurements of literacy as well as Standard Assessment Test scores (SATs) and GCSE results.

Research Questions:

The overall aim of this study will be to identify the role of language as a protective factor in children's educational progress.

Particular questions include:

* How closely related are a child's social and communication environments?

* How do components of a child's social and communication environments impact upon their language and literacy skills?

* How do social and child-based factors interact with a child's language and literacy to impact upon their social and educational outcomes?

* Are there protective factors that facilitate positive educational outcomes for the children.

Analytical strategy

We will undertake a preliminary descriptive stage to identify and provide means, ranges and standard deviations for all the key variables. This will include children's social and educational outcomes such as their SATs, GCSE results, friendships and social activity.

Following that, the first analytical stage is to build a number of composite variables that will also include a time element: a) a social composite (following Schoon et al, 2004) to represent the social resources available to the child and family including parent's education and employment, housing situation and other economic indicators; b) a communication environment composite comprising activities and resources available that indicate levels of interaction with the child such as book reading, television watching, parent-child interaction measures; c) a number of 'protective factors' - some of these will comprise a single measure such as the child's temperament, intellectual development, and others will be constructed to provide further composite measures such as attendance at child care and nursery facilities.

The second stage will be to model the child's developing language and literacy from early prelinguistic phases including babbling and attention to sound through to comprehension of paragraphs and vocabulary at the age of 8 years and scores on the Children's Communication Checklist at 9 years. Using latent class analyses and a series of regression analyses we will develop language and literacy trajectories.

Thirdly, we will undertake a series of multiple linear/logistic regression analyses to investigate interaction between the various composite variables. Structural equation modeling will be used to investigate possible mechanisms and the influence of factors over time. In particular the aim is to identify protective factors which result in positive educational outcomes for the children.