I tweeted yesterday that I am through with multiple choice tests, and I meant it too.
The problem is that I teach grade six in Alberta where students are expected to perform Provincial Achievement Tests at the end of the year which are, you guessed it, multiple choice. So I am left with a problem.
How do I, a flipclass enthusiast interested in valuing formative, authentic assessment also train my students how to have success in a multiple choice format?
The trite answer is to believe that students with substantial knowledge of the subject will demonstrate that knowledge in any format. Hey, if they know it, they know it, right? Of course, as practical experience shows, this is folly. My students, who show great ingenuity in solving problems and displaying knowledge performed poorly on MC tests all year. In fact, one girl who has been working hard and creating pages of documentation to explain her knowledge in a deep and rich way, scored 65% on the summative MC exam. This is due, in my opinion, to the multiple choice format which needs to ensure there is at least two correct answers, the right and the righter. When I have kids who are looking at the whole concept and identifying many facets to an issue, to hive off one is too simplistic. Thus, the right answer could be argued to be as correct as the righter answer.
So, I am left with the problem that I need to teach kids how to do the multiple choice exams and I need to make it a part of my practise during the year to ensure they get familiar with it, as opposed to cramming a 'how to write MC exams' unit in at the end of the year.
Which means I need to create some good, robust assessments of content in a multiple choice format. And I say I need to create them, because then I can avoid the problem of having questions from another teacher's point of view and not my own.
I have a second problem. I am not a sequential, logical thinker. The tests I make are error-filled and demonstrate the success of failure.
Sigh.
"Blink" and they'll miss it... Reading Gladwell illustrated a point I had previously missed: why do the "unstudied" do particularly well on MC alongside the "prodigies" and those in between seem to suffer the "stupid errors"?
ReplyDeleteAccording to instinct we are predesigned to introduce our own misconceptions as we struggle to incorporate the complete truth biasing a MC test toward those who only retain "common sense" and those who have become "experts".
As a summative assessment then it rewards only those who have not done any work -- in theory the experts were so before they arrived, and the unstudied remained blissfully ignorant.