Planning for grading for growth: Building the structure

Are you on board to use alternative grading? Here's what comes next.

Photo by Jake Hills on Unsplash

Don’t look now, but July begins this week, which means it’s almost August. And you know what that means: Whatever you’re thinking about doing in the Fall, it’s time to put some concrete plans to it.

David and I hope that alternative grading methods are part of the package of what you’re considering! In my last post, I gave a litany of questions to ask about courses in which you plan on doing alternative grading. By answering at least a subset of those (especially the Why? questions), you are well on your way to a successful launch of an alternative grading system.

So what’s next, if you’re still committed to trying something new with grading and you’ve answered all the questions? That’s what this post is about. I will outline some steps to start building a prototype alternative grading system in one of your classes, as early as this fall.

Before we get into it, I’m assuming that you have

• Picked out a course to try alternative grading in;

• Answered at least a handful of the questions from last time about that course;

• Looked at your calendar and scheduled some time and space to work on the items below. If not, then block out some time, because these require dedicated, focused thought; and

• Recalled the Prime Directive of Alternative Grading: Keep it simple. Pin this thought to the front of your mind, write it on a sticky note and keep it in front of you, or both.

By going through “Step 0” from the last post, you will be clear on the purpose and logistics of the course, why you are teaching it, and why you are going to try an alternative grading system in it. It’s now time to come down a few thousand feet in altitude and think about the specifics of the course.

Dream big with the course-level objectives

First, write out the overarching course-level learning objectives. These are the high-level aspirational goals for the course that spell out, in very general terms, what students should learn or acquire through the course.

There is no need to be terribly concrete here. In fact, many times course-level objectives are quite hypothetical. For example, Students will understand the concept of hypothesis testing or Students will gain an appreciation of the historical importance of the United Nations are acceptable styles for these high-level objectives. Some of these may be boilerplate from your department, university, or even your government and you may have little or no creative input on them. Whatever the case, write them out, because they will go in the syllabus.

Course-level learning objectives are definitely not the same kind of learning outcomes that I wrote about at length in the past. There, we emphasized that learning outcomes need to be actionable and observable, using certain kinds of verbs and a form like “I can <action verb>…<conditions>”. Course level objectives usually violate those rules —and that’s OK.

For example, here are a few of the course-level objectives from my recent discrete structures course. (The syllabus shows them all.)

Upon completion of MTH 225, you will be able to:

• Formulate and solve complex counting problems using computational thinking and the tools of combinatorics.

• Write clear, correct, and convincing arguments to explain the correctness of a solution using combinatorial proof and mathematical induction.

• Apply effective problem-solving skills in solving computational problems.

• Self-assess one's work and apply feedback from others to make improvements in that work.

These aren’t as pie-in-the-sky as they could possibly be — too much motivational-poster language is a turn-off for computer science majors — but they’re not exactly the action-oriented, highly specific outcomes we see with standards. (What is an “effective problem-solving skill” and how do I know if a student is “applying” it satisfactorily?) And again, that’s OK at this level.

Make the course more manageable with modules

Second, determine how the course will be subdivided into modules during your academic term. Most courses move through distinct stages as time passes. We refer to these stages as modules. A module might correspond to a unit of content. But modules don’t have to correspond to content at all. A module might instead refer to:

• A phase of a process. For example, if you are teaching a course that consists of a single semester-long research project, content is not really an issue — it’s all about the end result. You might structure the class so that “Module 1” is a period in which students do initial reading and formulate their research questions; and in “Module 2” student perform a literature review; and so on until “Module n” (the last module) in which students present their final product.

• Themes. For example, if you are teaching a writing class where the purpose is to have students do different kinds of writing throughout the semester, content might not be an issue here (because maybe students can pick their own topics, or because you don’t plan to assess mastery of any particular body of knowledge). Instead, there could be one module for each style of writing.

It’s definitely possible to overdo the concept of “subdivision” so that your course becomes a mindless assembly line, rather than a transformative learning experience. But at this level, where we’re breaking the course into a handful of logical units, we aren’t crossing the line in my view. What subdivision into modules does, is make the course easier to manage for everyone. Studies have shown that structure in a course helps students get and stay deeply engaged in a course¹ , and that’s part of what we’re after here.

A course taught during a 15-week semester will typically involve around 4-8 different modules. How you arrange your modules is up to you. But we recommend keeping the number of them in this “medium” range — say, no more than one module every two weeks, if you are thinking about their duration. Having too few modules (for example, just two or three for the whole course) defeats the purpose of subdividing the course. Having too many (for example, 20 or more, or possibly even one a week) makes each module so small that the package of assessment and grading may be difficult to manage.

Modules not only provide helpful structure for students, but they also help you think about how assessments might work and what grading approach might work best. A module might include a package of assessments that appear in each module. For example, each module might have a set of readings assessed with a quiz, followed by a discussion prompt for the class discussion board, followed by a unit test or a paper. That recurring structure can be good for students because it makes the workload predictable, lowering extraneous cognitive load. On the other hand, each module might have its own unique set of assessments. In the example above of a course that consists of one big project split up into phases, each phase likely has different requirements and standards. So, while there may be cross-cutting assessments (e.g. a weekly journal entry no matter what phase you’re in), it would make sense for each phase to have different assessments. (This makes each module more like a bundle in the language of specifications grading.)

Focus on specifics with the standards

Third and finally, we have one of the most important and difficult steps in this entire process: You’ll next make a first pass at specific, lower-level learning objectives that will be your “standards” for the course, and that you will eventually assess. These are different than the high-level, aspirational course-level objectives you wrote out earlier. Here, we want to identify the assessable units of knowledge or skill that students will acquire in your course and phrase them in a way that makes them simple to assess. In other words, at this point we are going to make a rough draft of our Clearly Defined Standards as given in the Four Pillars model.

We’re asking you to create these objectives even if you’re not planning to use Standards-Based Grading (in fact, you might have noticed that we haven’t even talked about what kind of assessment system you’re using yet — you’ll see why soon). So why write objectives now? It’s because they will form the core of any kind of assessment system (even traditional ones!). If, for example, you’re going to use specifications grading, you’ll use these objectives to guide and focus the process of writing the specs for each assignment. If you’re going to use ungrading, you and your students still need to know what things matter in your course. And of course, if you’re planning to focus on Standards-Based Grading, these objectives will be exactly what you assess.

The process of writing good standards is like a craft, somewhere between “art” and “science”. This process is so important that it got its own post back in February. To summarize that article:

• A standard is a clear and measurable action that a learner can take to demonstrate their learning of some important topic or concept.

• We can identify three levels of standards: course-level, lesson- or module-level, and assessment-level.

• What are the differences between lesson- and assessment-level objectives? If you went block-by-block through your course and wrote down every single skill, fact, or concept that a student could possibly learn, you would have the lesson-level objectives — a very long list of extremely fine-grained learning outcomes. If you stopped there and made those your “standards” — thereby committing yourself to assessing and reassessing each of them, for your entire student audience, over and over again, for the whole course — you’d need a long quiet lie-down after the semester. Instead, you can go through that list and consolidate some of those into thematically linked groups of objectives, and then assess the group. If you do that, then the related clumps of objectives are your assessment-level objectives, and those will be your “standards”.

• Both lesson- and assessment-level objectives need to be clear to the student and observable by you, because otherwise you may not be able to assess them, or what you assess may not be what you think you are assessing. Bloom’s Taxonomy is a handy guide for picking good verbs for these.

What that earlier post didn’t get into, is the process of actually creating this list. The way I do it is like this:

1. Go through the entire course, one module and one lesson at a time at the lowest possible “altitude”, and write an exhaustive list of all the things that students can or will learn in my course. Here is the result from the planning stage of my discrete structures course. This has 58 (if I counted right) lesson-level objectives on them, although many could be broken down even further into smaller topics.

2. Then go back through that list and group lesson-level objectives into clusters that I can assess. Here’s the result, in which that previous list of 58 has been distilled down into 20. (A roughly 3:1 ratio of lesson-level to assessment-level objectives has been pretty consistent in my experience.) The assessment-level objectives involve grouping thematically similar lesson-level objectives together; for example my Learning Target 1 (I can represent an integer in base 2, 8, 10, and 16 and represent a negative integer in base 2 using two's complement notation) is a combination of the first three items on the lesson-level list. There are also some lesson-level objectives that just didn’t make it onto the list of assessment-level objectives (like the one about stating the Division Algorithm). These represent things that students will encounter and which I hope they learn, but they don’t rise to the level of importance in this context to merit their own assessment.

In other words, you can’t do it all, so you have to decide which topics or concepts are truly worthy of the time and energy of repeated assessment, and which ones are just cool things to learn. (If you have topics that are neither… cut them from the course altogether!)

My process of going to the lowest possible level first, writing down a massive list of extremely fine-grained objectives, then working your way back up to the assessment level may not work for some people. For example, you might not be very familiar with the course (because it’s a new prep, you’re using a new text, the course was redesigned, etc.) and therefore may still figuring out what the students are going to be learning on the lesson level. If that’s you, then I’d suggest starting at the lowest level you feel comfortable with now and going through the process above. You can always come back and update your lists once you have gotten more familiar with the course².

What’s next

Suspiciously absent from the process above and the “Step 0” in the previous post is the actual choice of assessments and the grading system itself. Postponing the choice of the grading system is intentional, and it may feel to you like being asked to eat your broccoli before you can have ice cream³. Picking the grading system is the fun part, right? Why put it off?

Well, designing a grading system is fun, in fact it’s so fun that some people make the mistake of starting there, and then building the course around the grading system. But this is no better than some approaches to course “design” (and I use that term loosely) that we find in traditional grading systems. You know the ones; the professor starts with “three tests and a final exam” and the table of grade percentages, then arranges the content and lectures to drive toward the tests and their grades. The main problem with that approach is that the humans in the class are not involved the design process. A system designed for humans, but without those humans in the loop during the design process, is a blueprint for disengagement.

We are all about student growth on the other hand. And so, armed with a sound grasp of the course background and its human components, as well as how the overall story of the course should unfold, we’re equipped to think about how the grading will work. That’s next.

---

Thanks for reading. What’s your take on these steps? Do you have a different process you like to follow? Let us know in the comments.

1. The study that’s linked here refers to online courses, but (1) my experience suggests the same results hold for all modalities, and (2) the lines between modalities are so blurry these days that a result in one space is definitely worth considering for another.

2. But as a reminder — it’s almost July. Probably time to get serious about familiarizing yourself with the low-level objectives.

3. I happen to really like broccoli, more than some ice creams, which might partially explain why I’ve written everything in this order so far.