12,000

We have over 12,000 students, from over 100 countries, within one of the safest campuses in the UK

93%

93% of Lancaster students go into work or further study within six months of graduating

Home > Research > Projects > Variability in child language
View graph of relations

« Back

Variability in child language

Project: Non-funded ProjectProjects

1/04/07 → …

A feasibility and pilot study on the exploitation of the Child Language Survey

This project is a feasibility and pilot study on the exploitation of the Child Language Survey. It is led by a cross-faculty team including Katie Alcock and Kate Cain (Psychology), Andrew Hardie and Sebastian Hoffmann (LAEL) and Paul Rayson (Computing). The RA on the project is Nicola Pooley (Psychology/LAEL).

Background: The Child Language Survey

In the 1960s, the Nuffield Foundation funded the Child Language Survey (CLS), a project which gathered a vast collection of data on child language from the ages of about 8 to about 15.

Consisting of transcripts of child language, both written and spoken, collected from a number of schools around the UK, this data was published in the late 1960s. Its extent has been estimated as a million words (of which 80% was spoken, 20% written).

While some university libraries possess copies of the transcript booklets, the CLS has long been unexploited, despite its potential value, because it is not in the digital format crucial to modern large-scale text analysis.

Research programme

In this pilot project, we will digitise a selection of the CLS data (both spoken and written, in the former case including audio recordings)

create a comparable modern-day sample of data from the same or equivalent schools in London and Leeds

investigate the use of this data in studying children's linguistic variability

We will investigate the evidence in the data for the following three skills on the part of the children:

Planning the text: We will compare the coherence and cohesion of written narratives. These measures will also be analysed in relation to measures of fluency (text generation measures).

Generating the content: We will consider meaning-based dimensions of language (vocabulary and the development of ideas) and rule-based dimensions (sentence structure) within each modality. Complexity and diversity will be examined.

Transcribing ideas into written language: Spelling ability and writing conventions will be analysed. Children can also appear to be good or poor spellers by their choice of easy or hard words to spell in their written compositions, so spelling success will also be measurable in terms of written word frequency and length.