Standardized Learning
One conclusion I’ve come to after years teaching – testing and assessment are poorly used as a way for students to learn.
This is curious and unfortunate because students for the most part DO get motivated and energized through tests and quizzes. The pickle is, the way they are designed doesn’t make the test a learning experience and rather is meant to trick students. I’m calling for all teachers to review the way they test and I’m offering one example using the popular convention of testing – multiple choice questions.
I recently began one of my classes by writing the following on the board. A typical, 3 truths / 1 lie activity where students try to guess the lie.
This year I resolve to ….
1. grow my hair long
2. plan my classes better
3. travel the world and teach
4. get a new coffee maker
It’s a great activity for teachers to share themselves and also for students to do and allow the teacher to get to know them. However, I’m teaching teachers so I took this opportunity to go beyond the activity and ask them what this multiple choice question might say about assessment and how we decide / design these questions.
What’s remarkable about this question is that you can pose it two ways. One – which statement is the lie? Two – which 3 statements are the truth? Now you might think this is just semantics but I believe if we created multiple choice, standardized assessments where the students were asked to not choose just one right answer but three right answers – they’d learn a lot more. They’d be encountering a lot of “right” knowledge and not trying to side step through a labyrinth of wrong.
Here’s another example.
A typical standardized multiple choice question for language students might be;
Beth ___________ to the store every day.
a) has b) is c) went d) liked
A multiple choice test that would actually give students more success and help them learn would be them choosing the 3 appropriate language forms.
Beth ________ to the store every day.
a) went b) likes c) goes d) has gone
It’s important that students choose 3 right answers and not be asked to choose the 1 wrong answer. This way, we can give marks for right answers. This way they feel “success”.
This is just one of many ways we could rethink assessment and make it more about “learning” and less about tricking students. Do you have any other ways?
PS. The 3 correct resolutions for this year are 2,3,4!
Interesting post. I like the idea of setting ourselves the challenge of writing MCQs that suck less. To be honest, though, I am not sure that the 3-right-answers idea makes them less sucky. Obviously, if I give you an MCQ with 4 options and tell you that 3 are correct it is easier to choose one of the correct answers. But if you have to find all 3 of the right answers to get the mark, the new MCQ is no easier than the traditional one. You still have to go through each choice and ask yourself “Is it right or is it wrong?”
Another problem: One of the really sucky things about the sort of MCQs you have here is that they use little sentences that have absolutely no context. Now, if there is only one right answer, it would be easy to make the sentence longer or add other sentences to enrich the context. Your suggestion, however, will tend to force the exam writer to decontextualise as much as possible.
You say at the end that you don’t want exams that trick students. I agree that this is the way things ought to be. Yet, if we put ourselves in the shoes of the examiners, we see that there is a certain rationale behind the practice of trying to catch students out. In the ECPE, for instance, there are only 40 vocab MCQs. In a good exam (where is it?) students would be asked to show what they know. In the real exams they have to show what they don’t know. Those who don’t know less, get the highest marks. In other words, exams like the Michigan ECPE are tests of ignorance, not of competence.
The challenge (but it would be too stupid) would be to design a better exam to test the richness of a student’s vocabulary (bearing in mind that this will be a student who has been studying English for 5, 6 or 7 years) using only 40 questions. My guess is that if we took up the challenge, we would eventually come to the conclusion that it would be better to just bin the MCQs and use something completely different that did actually allow students to show what they know.
Torn,
I agree with your 2 basic and well reasoned premises. 1. Use context more and as much as possible 2. Bin the tests if we can and have the power and use something that allows the student to show what they know.
However, it does get tricky when the emphasis is on “show what you know” instead of “show what you don’t know”. Knowledge is such a personal thing and that’s where standardization is truly not a good barometer of performance and I’d even suggest, rubrics and other performance evaluators are even just as bad as mulitple choice because they still make for too strict an interpretation about what is knowledge/knowing.
I guess we are in a dilema. But my main goal was to suggest we make tests where students can succeed much easier. Tests where students can learn something from the things in front of them, from the questions (and a proper question is the starting and ending of all knowledge).
Oh I forgot. The methodology of 3 right supposses that choosing 1 wrong will always give you wrong (0). Choosing a right will give you 1/3 of the mark. You don’t have to attempt / get all 3 right to get marks.
David