Standards or specifications?
A fundamental choice in alternative grading maybe isn't quite so fundamental
I’ve heard a fair amount of anxiety from new alternative graders who want to know “Should I be using standards-based grading or specifications grading?” This feels like a really important decision, one that sets you on a path that determines how you’ll design your whole grading system. I get a sense that folks who are new to alternative grading think that they must choose one or the other, and then follow the rules of that approach precisely or else they’ll fail.1
This is a false dichotomy. You don’t need to make an all-or-nothing choice between these two approaches to alternative grading, because they aren’t mutually exclusive. Today we’ll take a look at standards, specifications, their differences, their similarities, and how to decide what you should be using in your own classes.
What are standards?
Let’s start with this definition from my alternative grading glossary:
A standard is a clear and observable description of an action that a student can take to demonstrate their learning of some specific topic. … Often used interchangeably with terms like “objective”, “learning target”, etc., although many sources try to distinguish between these.
Here are some examples of standards from my own practice, from guest posts, and even from our book:
I can create sets using set builder notation and translate sets into this notation. (From my own Communicating in Mathematics class.)
I can write brief descriptions of specific time management techniques, including time-blocking and the Pomodoro technique. (From a freshman seminar class we invented for our book.)
Represent and describe a physical situation with a sketch and detailed labeled diagram (From Robynne Lock’s University Physics class, which you can find in our book.)
Create a model to illustrate the relationships among DNA, chromosome, gene, and allele. (From Jenni Momsen’s General Biology II class.)
A standard usually describes a single specific skill that a student can demonstrate on an assignment. Individual assignments might involve multiple standards – for example, one quiz might assess students on using several different methods that they’ve learned in recent classes. Their skill in using each method addresses a different standard. The same standard might appear on multiple assignments, as a way to show a student’s progress over time. As a result, you can record students’ work on each standard separately to get an understanding of which skills a student has learned, and which ones they still need to work on. Assignments that work well with standards include quizzes, exams, and problem sets – all situations where you might assess multiple separate skills.
Standards are usually fairly specific or narrow, but they can also be broader. “Process” or “method” standards, such as “I can choose an appropriate method to evaluate an integral”, are valid too.2
What are specifications?
Next let’s take a look at specifications:
Specifications are a detailed description of what a “successful submission” of an assignment involves. “Specs” can include mechanics like completion, word counts, spelling, grammar, etc., but often focus more on general qualities, integration, synthesis, etc.
The key here is that specifications focus on a whole product, not just an individual skill. Specifications, by their very nature, tend to be longer than standards, because they describe the overall qualities of a student’s work rather than individual skills used in creating that work. It’s critical that specifications don’t describe every possible quality of good work; rather they are meant to describe the truly essential qualities required to show appropriate learning.
There are lots of examples in our book (Chapter 6) as well as in case studies on this blog, so I’ll just give one quick example here. Below are some specifications from my own Communicating in Mathematics class, which is a discipline-based writing class in our math major. Students write proofs — detailed, multi-page logical arguments — that are assessed with this list of specifications:
Include a true and correctly worded theorem statement just before the proof.
At the very beginning of the proof, state all assumptions and explain what will be proved (even if this repeats what was in the theorem statement).
State the proof method being used. If using proof by contrapositive, contradiction, cases, or induction, clearly state the new assumptions or statement being used.
Give clear, correct, and organized justification for each statement. Clearly show which results justify each statement.
Have no important errors or omissions in mathematical reasoning or justification.
Define all variables.
Cite definitions clearly and correctly each time one is used.
Follow all instructions on the portfolio sheet, including additions or changes to these specifications (most often, this includes adding examples and an Overleaf link).
These individual items might look like standards, and that’s more or less true. But the key is that an assignment graded with specifications earns just one overall mark for meeting all of the specifications at once: Either a student has thoroughly “put it all together”, or they haven’t (yet!). I don’t keep a separate record of whether a student knows how to include a correct theorem statement, state all assumptions, and state the proof method being used. Rather, those are each important enough to the overall proof that if they don’t do any of those, the entire assignment earns a “not yet” grade.
If you’re thinking that this could be a bit harsh, many instructors add a bit of flexibility with specifications by requiring – for example – a student to meet all but one specification, or perhaps meet all of a list of “core” specifications and at least some “additional” specifications. Indeed, if you follow the link to my actual specifications, you’ll see that there are additional “excellent” level specifications that a student doesn’t have to meet in order to pass the assignment, but if they do, they can earn an “Excellent” mark.
As a result, specifications work well with assignments where you want to see a student’s ability to integrate multiple skills into one solid product. Essays, projects, portfolios, proofs, and other longer-form written items all work well with specifications.
There’s also some terminology trouble: Some authors have used the term “specifications” to essentially mean “a description of what’s in a standard”, and likewise others have used the term “standards” to describe long and detailed descriptions that are applied to an entire assignment – in other words, specifications! So, be careful when reading about these ideas, and focus on the ideas rather than the names that authors use.3
How do I choose between standards and specifications?
While I’ve been talking about standards and specifications, I haven’t been focusing on “standards-based grading” or “specifications grading”. That’s because I see standards and specifications as elements that can be mixed and matched within any grading system. They’re useful in different situations — some assignments work better with standards and some with specifications — so one class could include both standards and specification-based assignments.
It might help to think about the previous section this way: Standards focus at the level of individual skills. Specifications focus at the level of the whole assignment. This means that when you’re deciding to use standards or specifications, you should be thinking about your big-picture goals for a class, and how each assignment contributes to them.
So to decide which one you should use in a given situation, ask: “Do I want to focus on individual skills, or big-picture integration?” The answer is likely different for different assignments, and you might even want to build up from assessing individual skills on one assignment (a quiz), to a holistic synthesis of them on a later assignment (such as a project). If you’re not sure, this is a good time to rethink the goal of each assignment, and to refocus it in one direction or the other.4
As mentioned above, quizzes, exams, and problem sets often work best with standards. These are ways to contribute to an overall goal of building specific skills that might be needed in later assignments, future classes, certification exams, or employment. These assignments tend to involve questions at the lower levels of Bloom’s taxonomy, which is another pointer towards using standards. You can list the specific standard being assessed above each problem (or perhaps list all of them at the start of the assignment) and then record a student’s progress on each individual skill using a mark that indicates progress like “Successful” or “Needs new attempt” (standards-graded assignments often benefit from students making a wholly new attempt on a new problem that addresses the same skill, to show that they can use that skill “fresh”).
Likewise, longer-form work like essays, projects, portfolios, and proofs work well with specifications, because these types of assignments are all about putting together multiple ideas into one coherent demonstration of understanding. These assignments contribute to overall goals of synthesis, analysis, integration, and even creativity – in other words, the upper levels of Bloom’s taxonomy. You can list the specifications on the assignment, or as I do with proofs, post one overall list of specifications that apply to every assignment. Then assign one mark for the assignment, such as “Successful” or “Needs revision” or one of the levels in the EMRN rubric (specifications-graded assignments often benefit from revising previous work, rather than starting fresh).
As a result, saying that a class uses “standards-based grading” is really just a shorthand for “most or all of the assignments in this class are graded using standards”, while saying that it uses “specifications grading” is a shorthand for “most or all of the assignments in this class are graded using specifications.”
It’s entirely possible that you might teach a class where only standards or only specifications make sense. But you don’t have to – indeed, you shouldn’t – limit yourself to only standards, or only specifications. It’s much better to think carefully about what’s appropriate for your students, your class, and the goals of your assignments, and then mix and match as needed.
If you want to learn some ways to keep track of all of this in a gradebook and put it together into a final grade, check out my posts on “Using a gradebook with alternative grading”, part 1 (the gradebook) and part 2 (final grades).
What about ungrading and other options?
I’ve been focusing on standards and specifications in this post. You might be wondering about other alternative grading options like ungrading (whatever that means), contract grading, labor-based grading, and many other names that are in use.
Time to make a controversial claim: Most alternative grading systems are fundamentally based on standards and/or specifications, even ones like “ungrading”.
That’s because, in any class, there should be a clear description of how a final grade is determined. The elements involved in determining an “A”, “B”, if clearly stated, are standards or specifications.
For example, in an ungraded class where students create a final portfolio and use it to make an argument in favor of a specific grade, there are descriptions of what students should show or demonstrate in that portfolio in order to earn each specific grade — those are standards and specifications. A portfolio might need to include items that demonstrate a student’s understanding of specific skills from the class — those are standards. Or perhaps a portfolio needs to holistically show a student’s skill with professional writing — those are specifications. Perhaps there’s a mix, which is what I do in my ungraded Euclidean Geometry class, where final portfolios need to address both specific skills (geometric topics) and show overall growth in certain areas.
Likewise, contract-based grading systems use specifications: the contract spells out the specifications. Labor-based grading is similar, where the specifications are entirely focused on completing a certain number of assignments, regardless of quality.
This isn’t to say that all grading systems are just “standards” or “specifications”. They definitely are not! There are as many ways to mix, match, and add interesting additional elements as there are instructors. Each instructor brings their individual philosophy, their understanding of their discipline, their students, and their overall goals. All of those contribute to the huge and fascinating variety of alternative grading systems that are in use.
But in all of them, standards and specifications are fundamental ingredients that form the system’s foundation. Thinking about those options from the beginning can help you focus in on what’s most appropriate for your situation.
How do I get started?
It’s still early summer (indeed, some schools haven’t even finished their terms yet!). But if you’re thinking ahead to next year’s classes, we have some resources for you:
Chapter 11 of our book is a workbook that steps you through the process of designing an alternatively graded class. It takes this viewpoint that I use in this post: That fundamentally an alternative grading system is founded on a set of goals, and those goals can be expressed as a mix of standards or specifications.
A few years ago, Robert wrote a series of posts intended to help readers think through designing a new class, much like our workbook. Here’s the final post in that series: Planning for Grading for Growth, which includes links to each of the earlier posts.
There are tons of case studies and guest posts on this blog (and in our book) to help inspire you with ideas for designing your own class. Check them out! Here’s a link to just a few of our guest posts.
You don’t have to choose between standards or specifications. This isn’t an all-or-nothing choice. Rather, I hope you’ll think of them as tools that can help you design an appropriate grading system for your situation, your students, in your class. Don’t let the names limit you: You have the power to design a system that’s right for you!
Usually there’s a third choice in the list, “ungrading” — whatever that means. We’ll see how that fits in with standards and specifications later in this post.
This kind of “method” standard addresses a common question about standards-based grading: “Doesn’t writing the standard on each problem give away the right way to solve it?” There are two different possible skills here: One is actually applying a method correctly (“I can use integration by parts”), while a different skill is choosing one method from several (“I can choose an appropriate method to evaluate an integral”).
To add a little complexity, there’s another layer here: When grading with standards, there are often more detailed “success criteria” that you will use, and should communicate to students, to determine whether a standard is met successfully. These are essentially specifications in miniature. For example, what exactly does it mean to successfully create sets in set builder notation? Well it means defining all variables appropriately, identifying an appropriate predicate, placing the set elements in the correct part of the notation… these are exactly specifications, all of which a student needs to meet in order to pass the standard. But because those are all contributing to assessing a single skill, separately from others, I see this situation as truly involving a standard.
A reasonable question at this point is: “What about bigger projects/essays/etc. that require students to successfully use individual skills?” For example, coming from my math perspective, a written proof (assessed with specifications) often requires students to use many individual skills (assessed with standards). To me, this points to specifications, because your goal is still to see if students can put all of those individual skills together. Your specifications can include specific skills which must be used correctly within the bigger assignment, or even a general statement like “all mathematical arguments use course content correctly.”
Super-useful explanation of this distinction. And I appreciate the important connection made between standards/specs and ungrading.
This is definitely a useful explanation of the distinction between these two types and it gives us some conceptual tools to know when to apply either (or both!). However I can’t help but feeling that the most important portion (for me) of standards based/ specs based grading or all ungrading for that matter, is left out here, which is actionable feedback. I know this is not ignored in this blog and I’m sure it is throughly covered elsewhere (whether that be in the substack or the book I have yet to read) but I find that the greatest benefit of ungrading comes from the fact that students receive feedback that contributes to growth in ways that regular grading does not. Without clear detailed feedback, I find there’s no real distinction between “ungrading” and grading.