Archive for October, 2010

Flattery or feedback?

Thursday, October 28th, 2010

Reading this week’s New Scientist (23rd October 2010) on the train on my way to the Centre for Distance Education Conference in London yesterday, I found an interesting opinion piece from Clifford Nass ‘More than just a computer…’ The article talks about how people like to be flattered, even when the flattery is unjustified. This resonates with my own reaction to praise – I like to be told I’ve done well even when (and perhaps especially when) my own gut reaction is that I haven’t! (more…)

Multiple-choice questions – love them or hate them

Sunday, October 24th, 2010

I seem to have spent quite a lot of time muttering about multiple-choice questions. You might be wondering what I have against them. It’s partly that students don’t really have to engage with the assessment process in as deep a way if they are just picking an option rather than constructing a response for themselves, but also because it is possible to guess a correct answer – so when used summatively, if a student has got a multiple-choice question right, you don’t really know whether they understand a topic or have demonstrated a learning outcome. There are ways of dealing with this problem, primarily confidence-based marking.

Also, to be fair, I have seen some excellent and creative uses of multiple-choice questions. When the OpenCETL websites are mended I will give some examples. So if the e-assessment software you are using only allows multiple-choice questions, or if there isn’t another way of asking the question you want to ask, don’t despair.  Some multiple-choice questions are distinctly better than others!

Writing good interactive computer-marked assessment questions

Thursday, October 21st, 2010

I run a lot of workshops trying to help colleagues to write good e-assessment questions. There are usually lots of brilliant ideas in the workshop, but somehow we end up slipping back into using lots of multiple choice questions because people think they are reliable.

I suppose it is true that the answer matching is easier to set up for multiple choice and multiple response questions, but beware – just because you (as author) can easily identify one response as ‘correct’ and others as ‘incorrect’ it doesn’t mean that the question is behaving in the way you expect. The question might be ambiguous or there might actually be more than one correct option. Or – the most common problem – whilst some options are definitely correct and others are definitely incorrect, there may also be options which could either be correct or incorrect, depending on your interpretation of them. (more…)