Since returning from sabbatical in 2017, I’ve taught pretty much just physics (with one or two geo courses here and there), so my physics teaching game has been on my mind a lot. The main breakthrough I’ve made in my teaching over the past few years has been to adopt standards-based grading, which has allowed me to communicate my expectations to my students more clearly. I still struggle, though, with developing good standards for my particular course and my students – item 2 from Brian Frank’s list:
Easier said than done, because it’s super hard to actually have the three things it requires : (1) having an explicit idea about what it is you want students to do, (2) having that idea be a good idea, and (3) having a way of making that transparent and accessible to students.— Brian Frank (@brianwfrank) September 9, 2019
A major issue has been that the standards are basically a mix of a small, disconnected tasks that I expect the students to do – these are the easy-to-assess, easy to communicate ones – and big, “squishy”, higher-order skills I want students to develop. An example of the former is:
I can differentiate between isolated and non-isolated systems both conceptually and based on data about those systems.
I can calculate kinetic energy for individual objects and systems.
On the other hand, my list from last year had standards like:
I can reason about the motion of an object undergoing constant acceleration.
I know, “reason about” is a bad clause if you are going by Bloom’s taxonomy, but I find it hard to express the bigger picture of using what the Modeling Instruction folks call the constant acceleration (kinematical) model. Can you use the model to make predictions? There’s a lot you have to be able to do in order to get there, and a lot of gradations inherent in the word “use”: you could be using the ideas well in a qualitative sense, but not have the skills developed well enough to quantify your predictions. You could rely exclusively on memorized formulas without really knowing what they imply or where they come from, but use them effectively to make quantitative predictions. So I kept “reason about”.
I’ve also found that students may be able to succeed at enough of the standards to do well in the course, but still not be able to use the skills they’ve developed in an independent way. Basically, the course doesn’t do enough to challenge the traditionally successful students, and doesn’t allow the less traditional students enough say in pursuing problems that don’t fit well on tests. Students need a way to distinguish themselves that’s true to them, not just convenient for assessment.
So I’m trying to start the year off right by re-evaluating my standards in light of what I think is the most important idea I hope students get from the course: to slow down and reflect on their ideas in a methodical, systematic way. I’ve divided the course up into topics, each corresponding to a different skill or way of thinking: measurement, descriptive kinematics, momentum, forces, and energy; rotation and gravity, which we treat at the end, are outliers – more applications of ideas treated elsewhere in the class than new ideas on their own. I want to use an approach similar to Modeling Instruction (I’ve read quite a bit about MI, but I don’t feel like I understand it well enough to adapt it for a calculus-based university course), but focusing on exploring the following aspects of each topic:
- Making sense of experimental data
- Describing information using multiple representations
- Building a model and using it to reason about situations
- Applying mathematical, logical, and communication skills
- Reflecting on learning
I’m looking at these as if they are “folders” in which I can put my existing standards. Some of them take the place of the squishy, big-picture standards I used to have.
The advantage of this arrangement is that I can also then have a standard in each aspect/folder that asks students to do something distinctive – something that is theirs – that I can point to as a success beyond just quiz and homework questions:
- Making sense of experimental data – I can develop my own comparisons between data and predictions from a model or simulation
- Describing information using multiple representations – I can choose and translate fluently among the most appropriate representations of a situation
- Building a model and using it to reason about situations – I can propose and solve significant problems using reasoning based on the unit’s main idea (or: problems that incorporate more than one unit’s ideas)
- Applying mathematical, logical, and communication skills – I can independently identify situations in which significant mathematical reasoning or skill is needed, and use those skills competently or I can express complex physics ideas effectively in written or graphical communication
- Reflecting on learning – I can test or otherwise identify the limits or assumptions of models or I can thoughtfully express changes in my own thinking about physics
I’m planning to grade the “small” standards on a 1-3 basis (1: misses the point; 2: getting there; 3: meets standard), and these “big” standards on a 1-4 basis (4: distinction). A student’s grade in each topic will be the highest grade in each of the aspects. So, for example, if a student has a 4 in model building for forces, they get a 4 for the force section of the course. I’d love to require students to try for distinction in more than one aspect of a topic, but I’m not sure how to communicate that to students on Canvas (i.e. grade it) – and if I can’t communicate something, then the standards-based approach loses its luster.
So, some questions for you:
- How can I use standards to signal to students that I want them to step back and think about what they’re doing in a methodical way? (To confront their expectations, biases, preconceptions…)
- What makes a “significant” standard in your experience?
- Have you had any experience with standards-based grading in an intro, calculus-based course at the university level?
- Do you have any ideas about how to improve the system I’m proposing?