Published June 14, 2024 | Version v1
Conference paper Open

Investigating Automated Grading Techniques in Effec-tively Teaching Advanced Excel to Large Groups Of Ac-counting Students

  • 1. ROR icon University of Pretoria


  • 1. ROR icon University of Applied Sciences and Arts Northwestern Switzerland
  • 2. ROR icon University of Pretoria


The global student population is growing while faculty size remains constant. Lecturers shoulder an increasingly heavy teaching burden with con-comitant pressure to increase their research outputs. Automated marking can al-leviate the teaching load on lecturers, especially when the automated marking setup does not demand additional programming skills. This case study is an in-depth description of a course in advanced Excel for accounting students, and how it changed from paper-based assessment to fully online assessment. The move to automated assessment was motivated by large and increasing student numbers and limited lecturing resources, a phenomenon particularly salient in developing countries, where the case study is situated. An online assessment structure was developed by the lecturing team since the university does not have additional funds for costly automated grading tools. It used existing re-sources such as the university's learning management system and implemented simple string matching as opposed to more complex semantic similarity ap-proaches. We found the success of the automatic grading depended on two as-pects: the students adhering to a clearly defined process during assessments, and the Learning Management System’s (LMS's) offerings in terms of grading functionality. The LMS used in the automating grading system described here, offered only basic matching functions for grading but no text manipulation functions required to build more robust and general-purpose grading solutions. This highlights a possible gap for LMS providers to address. In this case study we also touch on lessons learned in automated marking, improving marking ef-ficiency, and managing the constraints of the available tools.



Files (265.2 kB)

Name Size Download all
265.2 kB Preview Download