Home / News / Academic Writing Analytics update

Academic Writing Analytics update

CIC has been developing AWA (Academic Writing Analytics), a rapid feedback tool for students at all levels (1st years to PhD) to help them reflect on how effectively they are using academic language in their writing. The Australian recently covered this development in a story about UTS (June 3rd).

This tool currently makes use of a natural language parser in development at Xerox research labs, with whom we have a longstanding research collaboration. This will be combined with other open source tools (and possibly products) into an integrated platform. AWA identifies hallmarks of educated, scholarly writing, as summarised below: skilful writers signal an important move or claim (black text) by using specific linguistic forms (purple) which are recognised by their discourse community:

Metadiscourse markers for academic writing

XIP had no interactive user interface, so CIC has been developing and testing designs to create a user experience that meets the needs of students and academics, as well as tuning the parser to different disciplinary needs.

It is important to emphasise that this is not automated grading but rapid formative feedback on draft texts. AWA is underpinned by research into the teaching of academic writing, is being co-designed with input from diverse faculties, and support units (HELPS and IML), and we are building the research evidence of its validity.

Examples of how AWA’s user interface highlights sentences in formal analytical academic writing:

AWA feedback 2

…and an example from autobiographical, reflective writing:

AWA feedback 1

AWA also provides a range of other visual analytics to provoke reflection about one’s writing:

AWA feedback 3

This blog post introduces some of the controversy that surrounds this fast emerging field, and the history to our work which started at the Open University in collaboration with Xerox research… Then watch a recent webcast introduces some recent work on reflective writing parser, developed with input from Rosalie Goldsmith in IML.

Leave a Reply

Your email address will not be published.