The biggest-ever archive of pupils’ schoolwork is being collected as part of a three year-project looking at how children use language as they get older.
Samples of writing from English, history and science lessons from pupils aged five to 16 will be investigated by the “Growth in Grammar” project, led by researchers at the University of Exeter.
Thousands of pieces of work will be collected from schools across the UK, before being collated, digitised and finally made available as a free online resource for other researchers and education professionals in late 2018.
Researchers hope the database will lead to a better understanding of the way children use grammar, to help support teachers and improve the design of the school curriculum.
Lead researcher Philip Durrant, Exeter’s senior lecturer in language education, explained: “We know with adults what features they change when addressing a formal audience or writing a story – what we don’t know is how children do that.”
The research will result in “the most extensive archive of authentic written work in schools to date”
Using literature from the 1930s and previous studies which measure markers of progress in text, the researchers aim to identify 700 key features to track student progress. One marker is when and how pupils cross words out on the page.
Teachers and families must give permission for the work to be used, and each sample will be anonymised to protect pupils’ identities.
The project began in August 2015 and received £317,843 of funding from the Economic and Social Research Council (ESRC), giving it the capacity to analyse five thousand pieces of work from schools.
So far researchers have collected 1,500 samples, with the study due to conclude in August 2018.
Durrant says the research will result in “the most extensive archive of authentic written work in schools to date”. Samples will be collected up until autumn this year, before analysis begins.
“We see this as a way of giving teachers and students greater awareness and control over the linguistic features that are critical for ensuring all students get to be the best writers they can be,” he told Schools Week.
Comparisons can be made with the British National Corpus (BNC), a 100-million-word collection of samples of written and spoken language showing how words and attitudes have shifted.
Full findings from the latest version of the study – the Spoken BNC2014 – will be released at the end of 2017, though early findings on words concerning education were released in 2015.
The early release revealed that the word “education” appears more frequently in conversation these days – appearing 42 times per million words, compared with only 26 times per million in the 1990s dataset.