At the Open University we use different variants of our iCMA questions. So, to take a very simple example, when one student receives the question ‘Evaluate 3 + 7’, another might receive the question ‘Evaluate ‘4 + 5’. In summative use, different variants limit the opportunities for plagiarism. In formative-only use, different variants provide students with extra opportunities for practice.
Some question authors will allow variables to be generated automatically, which offers the potential to create a very large number of different variants from a single stem. Perhaps that’s a good thing. But I am always concerned that
- variants generated in this way may not be of sufficiently similar difficulty (a topic to be discussed in future posts)
- some versions of questions may look unnatural or ‘ugly’ (e.g. ‘Evaluate 4 + -5’)
- In the event of queries, I want to know what version of a question each student has received, not just the template used to generate the question.