An Automatic Grading and Feedback System for E-Learning in Information Technology Education
Peng Li, Lee W. Toderick · 2015
Abstract An Automatic Grading and Feedback System for E-Learning in Information Technology EducationIn the past few years, new, e-learning, virtual hands-on labs have been deployed in theInformation and Computer Technology Program at our University. However, the studentssometimes do not receive timely feedback because it is extremely time-consuming for theinstructor to grade the hands-on exercises. The students usually do not have the opportunitiesto fix the mistakes they made in previous labs due to lack of timely feedback.These challenges were addressed by the creation of an automated grading and feedbacksystem, called BroadReach Extended, for locally designed hands-on exercises. In this project,grading scripts were developed for multiple labs in three courses. A secure, centralized serverwas set up to host the grading scripts, manage log files, and store grading results. After astudent completes lab exercise on a virtual machine (VM), she/he clicks a ‘Submit’ hyperlinkon a web browser to download and run the feedback/grading script. The script runs on theVM to collect and parse the lab results. Performance feedback and scores are displayedimmediately locally, and also logged on the centralized feedback/grading server. Anacademic integrity system ensures that only the computer connection assigned to that studentis used by that student. The entire system is protected by IPSec, an encryption mechanismsecuring end-to-end connectivity between the student and the assessment server.With the automatic grading and feedback system, students can submit their lab workelectronically, review the feedback on their web browsers immediately, and correct mistakestimely. A graded lab experiment may be resubmitted and re-graded for a higher score. Manystudents are motivated to find and fix their mistakes. The grading data can also be used bythe instructors for assessment purposes.We believe this project can help improve teaching and learning in that 1) it provides studentswith multiple “trial and failure” opportunities to enhance hands-on skills; 2) it does notrequire major changes to the curriculum and 3) the solution is scalable and can be adjustedand deployed in other hands-on, e-learning courses in educational institutions.This paper has two parts. The first part will describe BroadReach Extended for facultyinterested in developing their own automated feedback and grading system. The second partwill provide quantitative statistics on student performance that demonstrate students dobetter when actively learning and afforded the opportunity to continue lab work untilachieving the maximum grade for that lab.