Copy the page URI to the clipboard
Jordan, Sally; Jordan, Helen and Jordan, Richard
(2012).
URL: http://journals.sfu.ca/ijea/index.php/journal/arti...
Abstract
Different variants of interactive computer-marked questions are used to reduce opportunities for plagiarism, and tools have been developed to indicate when variants are of significantly different difficulty. The tools include the facility index for each variant, a plot of the different scoring patterns for the variants and a probability that any difference could have occurred by chance The paper includes an example of the tools in use and identifies factors that have been found to lead to variants of different difficulty. An investigation into the impact on students’ overall scores is then described, with the conclusion that, for most modules, the effect is very small (typically ±0.5%). Factors leading to a larger effect include: a larger number of individual questions with significantly different variant difficulty; a smaller number of interactive computer-marked questions; and a higher weighting of the e-assessment component. Monitoring the behaviour of variants can lead to improvements in the quality of individual questions, better assessment design and enhanced insight into students’ misunderstandings.