Conference time: -
REAP Conference Fora (in programme order)
Forums > Raising students' meta-cognition Session > Gardner-Medwin & Curtin Case study
Subject: BAD PRACTICE to include a "don't know" option

You are not authorized to post a reply.   
Author Messages  
Tony Gardner-Medwin
Posts: 14
31/05/2007 17:48  
From Jane MacKenzie: Hi Tony, I'll pick up one of your provocative thoughts. You say: "4. It is (common) BAD PRACTICE to include a 'Don't Know' option with T/F or Best-Option Qs." Could you elaborate cos I just don't see why?
................

The implication of providing a don't know option is that students should use this option if their uncertainty is above some level. This is practically never sensible! It can only be rational for the student to omit a guess if the average penalty for a guess wold be worse than for a blank reply. This degree of negative marking is seldom deemed acceptable, and would put stress on candidates: Am I really so uncertain I should tick "don't know"?

To include "don't know" can actually selectively disadvantage the more able students - because they are the ones who have more insight into which answers are uncertain, and they could therefore be persuaded to omit uncertain answers, either because teachers exhort them to do so, or just by the implication of the presence of the "don't know" box - that ticking must sometimes be sensible. The same effect may also disadvantage the gullible or more deferential student who may be more inclined to follow teachers' advice, even if they have a hunch it's a bit dubious. Unless there is a very severe level of negative marking (for example, you could justify it if you use +1/-1 on best of 5 MCQs, where a guess would give you an average of -0.6 marks) students should always expect to gain by answering. It is never rational to believe that your answer is less likely to be right than a complete guess, so unless a guess would penalise you more than a "don't know", it is never rational to use the "don't know".

CBM of course does include strong (double) negative marking if you choose C=3. If a student is uncertain about an answer, they have the option to downgrade or avoid this negative marking by acknowledging uncertainty. Saying C=1 is equivalent to saying "I'm pretty unsure" (I think there is less than 2/3 probability I'm getting it right). It is worth answering at C=1 even if you think the answer is a near or total guess - though you're not going to pass an exam if all your answers are in this category.

Hope this makes sense, and thanks again.
Tony
You are not authorized to post a reply.  
Forums > Raising students' meta-cognition Session > Gardner-Medwin & Curtin Case study > BAD PRACTICE to include a "don't know" option


ActiveForums 3.6