Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.667615
Title: Developing predictive analytics to enhance learning and teaching in lab based courses
Author: Akhtar, Syed A.
ISNI:       0000 0004 5361 740X
Awarding Body: University of Surrey
Current Institution: University of Surrey
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Restricted access.
Access from Institution:
Abstract:
Computer-Aided Design (CAD) is a fundamental engineering course in universities, particularly for mechanical engineering students. It challenges students to bring together theoretical understanding, design and drawing skills. It is most often in the practical lab-based classes that students will seek help, and tutors typically approach the students to answer queries during practice sessions. To monitor students’ progress, tutors move around the classroom and observe the performance of the students. This requirement puts extensive pressure on the teaching staff to ensure students achieve the intended learning outcomes and have a good learning experience. Research to date has primarily focused on the use of interactive teaching content, practice tools and virtual learning environments. Little work has been done to improve CAD teaching using synchronous blended learning in lab-based classroom settings. In this thesis, an interactive CAD teaching system has been designed and developed to enhance the learning experience, and mitigate the pressure on often limited teaching resources. The system provided a two-way communication flow between student and tutor. It enabled tutors to share their computer screens with online students during lectures and then monitor their progress during practice sessions. Students were able to submit their questions or to request help as needed. In response, the teaching team could view and/or take control of the student machine and answer their query online using full duplex voice communication. The student experience of the system was collected using survey forms and focus group interviews. The system also recorded learning related behaviours including student login and logout times, and workstation location. This information was processed to understand their seating preferences and learning groups. It was then mapped on their final scores for application of learning analytics. This study followed three iterative cycles from 2011 to 2013 with first year undergraduate students from the MES department at the University of Surrey. The surveys received an excellent response rate of 70%. Analysis of the surveys and interviews suggested that students felt the system had a positive impact on their learning experience, with more than 50% of students positively expressing their willingness to reuse the system. Analysis of the learning behaviour data showed significant correlation between time spent in the classroom, duration of collaborative learning, seating position and the final learning outcomes. This correlation led to the development of a predictive model using linear regression. The model was integrated into the system to identify students who were potentially at risk of failing and therefore allow tutors to intervene in good time. Technically, the system was successfully deployed with an uptime, in terms of teaching sessions, of 98.3%. The study resulted in a stable interactive CAD teaching system that was used in a blended manner to facilitate face-to-face teaching and identify at-risk students. The study has contributed to our understanding of the value of passive data capture in learning analytics and suggests that this is a direction that warrants further research.
Supervisor: Xu, Wei; Warburton, Steven Sponsor: University of Surrey
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.667615  DOI: Not available
Share: