An analysis of the difference between traditional and computer-based assessment of IT-skills

Assessment performed by human examiners of candidates taking IT examinations is prone to error which, in some cases, has led to incorrect assessment results. One approach to reducing human error is to use computer-based assessment methods but these are also prone to error, albeit of different types, since computer-based methods have different strengths and weaknesses compared to human examiners. This paper compares the results of the computer-based assessment and human markers on sets of candidates sitting spreadsheet and database examinations as part of the pilot study for computerising a series of professional IT examinations.