Determining Optimal Blocking Sizes for Large Exam Groups When Data Matching

George Anderson, Audrey Masizana-Katongo, Dimane Mpoeleng

Research output: Contribution to journalArticlepeer-review


Computer literacy courses in the University of Botswana have a large number of students registered for them every year. Exams are multiple-choice, involving shading of identification details and answers on answer forms. This creates a data matching problem, since sometimes identification details are incorrectly shaded. In this paper, we explore the best way to structure such exams such that data matching can be used to match answer forms with students in the master registration list, in such a way that data matching accuracy is very good, due to blocking by exam group, and at the same time maximizing the class room size to save exam administration costs. Our results indicate the effectiveness of an optimal exam room group.
Original languageEnglish
Pages (from-to)80-86
Number of pages7
JournalInternational Journal on Information Technology
Issue number3
Publication statusPublished - 2014


Dive into the research topics of 'Determining Optimal Blocking Sizes for Large Exam Groups When Data Matching'. Together they form a unique fingerprint.

Cite this