Please use this identifier to cite or link to this item:
Title: Applying computer-assisted assessment to auto-generating feedback on project proposals
Authors: Al-Yazeedi, Fatema
Advisors: Payne, A
Keywords: Text mining;Ac1;Usability;Accuracy;Effectiveness
Issue Date: 2016
Publisher: Brunel University London
Abstract: Through different learning portals, computer-assisted assessment (CAA) tools have improved considerably over the past few decades. In a CAA community, these tools are categorised into types of questions, types of testing, and types of assessment. Most of these provide the assessment of multiple-choice questions, true and false questions, or matching questions. Other CAA tools evaluate short and long essay questions, each of which different grading methods and techniques in terms of style and content have. However, due to the complexity involved in analysing free text writing, the development and evaluation of accurate, easy to use, and effective tools is questionable. This research proposes a new contextual framework as a novel approach to the investigation of a new CAA tool which auto-generates feedback on project proposals. This research follows a Design Science Research paradigm to achieve and evaluate the accuracy, ease of use, and effectiveness of the new tool in the computer science domain in higher education institutes. This is achieved in three interrelated cycles:(1) based on the existent literature on this topic and an exploratory study on the currently available approaches to the provision of feedback on final year project proposals, a proposed framework to auto-generate feedback on any electronically submitted coursework is constructed in order to gain a clear understanding on how such a CAA tool might work; (2) a contextual framework based on the proposed framework for final year project proposals is constructed by considering both the style and content of the free text and using different text mining techniques; and (3) the accuracy, easy to use, and effectiveness of the implemented web-based CAA application named Feedback Automated Tool (FEAT)is evaluated based on the contextual framework. This research applies CAA and text mining techniques to identify and model the key elements of the framework and its components in order to enable the development and evaluation of a novel CAA contextual framework which can be utilised for auto-generating accurate, easy to use, and effective feedback on final year project proposals.
Description: This thesis was submitted for the degree of Doctor of Philosophy and awarded by Brunel University London.
Appears in Collections:Computer Science
Dept of Computer Science Theses

Files in This Item:
File Description SizeFormat 
FulltextThesis.pdfFile embargoed until 24/8/20192.78 MBAdobe PDFView/Open

Items in BURA are protected by copyright, with all rights reserved, unless otherwise indicated.