Sunday, 2 November 2014

Rethinking Tests

There have been a few tweets lately that have made me pause and think about tests. Even in the course where I spiral through the curriculum and no longer have units, I still give tests. I will admit that I'm not sure how to evaluate without using tests given the time constraints in place and the rigidity of the department in which I teach. I would love to sit down with each student and ask questions to elicit what they understand and where they are struggling, but I don't know how to manage that during class time. I look at Alex Overwijk in awe as I know he does this successfully. 

The tests I give, even in my spiralled class, look fairly traditional, but the way I "administer" (that sounds so formal when, in fact, I spend the entire class running around from one student to another) is likely not. The students in this class, grade 10 applied, all get two classes to finish the test. I take the time pressure off for everyone. I also don't let anyone hand in a test without answering all the questions, which means that I prompt as needed. I write whatever I ask or say in a pink or orange pen on their test paper and take it into consideration when I am marking the test. This means that the students who say that they don't care or that they have had enough still have to do work. I don't accept them opting out, especially since I know that they all can be successful. (So why am I testing them?) I don't have the same flexibility in my other classes though...

(I own this shirt)

In my academic classes (MPM2D and MHF4U this semester) I like to do stations for review the day before a test. The questions come from last year's test. I know the old tests are out there, some students have them, others do not, so I feel that this helps level the playing field. They work on the questions in random pairs on big vertical whiteboards and show me only the answer (unless the question asks for a proof in which case they bring the whiteboard to me!). They get a sticker when their answer is correct. Stickers provide motivation in ways I still do not fully understand. Here's the thing - when they do not get the correct answer I try to give them some feedback that will help them find their error and fix it. For example, if they had to determine the sine equation given certain information, I would tell them that the amplitude and vertical translation were correct in their equation, but they needed to look at the phase shift and period again. If they came back and the period was still incorrect, I would ask more specific questions about the problem to help draw out what they know about period then have them work on the phase shift again. This process of letting students correct their work in real time seems incredibly valuable to me. So I wonder why we cannot do this during a test. I realize that the logistics of attempting to do this with 30 students in 75 minutes do sound challenging. But I take issue with the snapshot of learning that we get with tests. Have you had a student do poorly on a test despite knowing that they understood far more than they were able to show you on the test? I have. Over and over.

I have started caring less about what I am supposed to do and more about doing what I think is right. When students don't succeed the first time, I give them another opportunity to show me what they have learned. After the 2nd grade 10 academic math test this semester, one student in particular had trouble with several questions bringing this student's  overall level on the test below where I thought it should be on this material. I had the student come in at lunch and we talked about how to tackle a couple of the questions. We had a good discussion - I found out what really was not understood and what this student could do given how to start the question. I think this was such a positive step forward for both of us. The student cleared up some misunderstandings and built up confidence by being able to show me all the things that they could do, but hadn't. I learned more about that student's understanding and what they really knew and showed that the learning, not the test, was the important part.

With what have other HS math teachers replaced tests? I would love to hear options. I teach at a school where the pacing guide is king, so I have an uphill battle, but I think this is the next goal in improving how I teach. I need help and would love to hear your ideas. Thanks in advance.


  1. Mary - I hear you. I have a similar situation in Algebra 2; I can see that a little one-on-one time with each kid who struggles makes an enormous difference, but there isn't time each day, sometimes even each week, to do that in a large class. So test grades suffer. I did retakes of portions of one of my exams this year - students had to get their tests sign, and give me a specific explanation of how they prepared to do better the second time around. Many earned over 10 points back on tests, which was encouraging - my tip of the hat to SBG which I cannot do in my school.

    But when push comes to shove where I teach, the Regents grade is paramount. If I help too much during an assessment, am I fairly preparing the kids for that exam? They will be judged by that grade (as will I, and my school) so it behooves me to get them ready. I know getting them ready requires a mix of one-on-one assistance while learning, and a simulation of the actual exam. It's an issue I struggle with all the time now.

    I emphatically agree with your tweet "I think we really need to overhaul school. Trying to fix within current system is very tricky." The system, especially in the US, and NYC in particular, is too huge and too broken. I feel like I am putting on bandaids (the really little ones that you always have left at the bottom of the box). Thanks for the provocative post.

  2. Great post Mary. So here's my 2 cents. I think you're right on the money when you say that assessment of learning is a snapshot of that moment. We (kids, teachers, parents) can all relate to the fact that some days produce better snapshots than others. A quick look in any yearbook can confirm this. So that's why spiralling is such a genius approach. You can take many snapshots. I think your kids are getting that message when you tell them that you value learning more than what they got on a test. You're telling them: You "looked" like this in today's photo and let's look at using a bit of photoshop to change that`.
    Should we support kids in taking "selfies"? As you know, I'm not in classroom right now but I think that's where I'd invest time. Defining what success looks like as you are solving problems and using some kind of portfolio (electronic, pictures, color codes, proof card ???). Show me you know. If Ss can recognize in their work evidence of learning that's what I'd assess. Easily said from where I sit, I know. Thanks for letting me think along with you.

  3. This comment comes well after this post was made, but here it goes anyway:

    I value what you say about doing what feels right and valuing what you are "supposed" to do less. I found that as I was teaching Grade 9 Applied Math more and more, I cared less about what they could do in a certain amount of time and more about what they could do. Period.

    I have similar questions about tests and making assessments authentic and meaningful when teaching strategies change to be more problem-based rather than procedural.

    But I also have questions about how to record and track conversations and observations. You share a great anecdote of a conversation with a student who didn't do well on a portion of a test. What do you do to "record" or track that conversation so that informs your judgement of their achievement? I'm interested in finding meaningful ways of questioning and then tracking interactions with students so that assessment is not only based on a product-based test.