Applications of Automatic Writing Evaluation to Guide the Understanding of Learning and Teaching

Authors

  • Peter Vitartas
  • James Heath
  • Sarah Midford
  • Kok-Leong Ong
  • Damminda Alahakoon
  • Gillian Sullivan-Mort

DOI:

https://doi.org/10.14742/apubs.2016.798

Keywords:

Learning systems, feedback, student writing, assignment scoring, large class management

Abstract

This paper provides an overview of tools and approaches to Automated Writing Evaluation (AWE). It provides a summary of the two emerging disciplines in learning analytics then outlines two approaches used in text analytics. A number of tools currently available for AWE are discussed and the issues of validity and reliability of AWE tools examined. We then provide details of three areas where the future direction for AWE look promising and have been identified in the literature. These areas include opportunities for large-scale marking, their use in MOOCs and in formative feedback for students. We introduce a fourth opportunity previously not widely canvased; where learning analytics can be used to guide teachers' insights to provide assistance to students based on an analysis of the assignment corpus and to support moderation between markers. We conclude with brief details of a project exploring these insights being undertaken at an Australian institution.

Downloads

Published

2016-11-25

Issue

Section

ASCILITE Conference - Full Papers