Standards and academic rigor are two challenging topics that I keep coming back to in these posts. On the one hand, it is exciting to hear people quote my statement, “Broader standards, not lower standards.” On the other hand, I recognize the incredible challenge of implementing standards that are both truly broad and high! This is why I was very excited to attend a two-day conference on alternative grading models this summer. (Shout out to our very own Renee Link, Professor of Teaching, Chemistry, who presented at the conference!)
The key takeaway of the conference was that alternative grading models provide a concrete framework for improving academic rigor in ways that are inclusive. However, the conference also made clear these methods are nuanced and require patience to explain and understand. In this post, I will attempt to introduce some of the ideas behind specifications grading and elaborate on how this grading system can work in practice. If what I say makes no sense to you, keep in mind that we have experts right here on campus that can offer a clearer picture! Also, watch for upcoming events from the Division of Teaching Excellence and Innovation if these approaches intrigue you and you are interested in learning more.
Before we jump in, I did want to offer a few disclaimers. First, one challenge of explaining specifications grading is the fact that it depends on multiple interacting pieces. You need to understand the various pieces before it makes sense. Focusing on only one piece can create a sense that this will not work because it is unfair to students, too much work, does not scale, etc. I also want to acknowledge that specifications grading is not an approach that works for every class. That being said, I discovered through the conference that it applies to more classes than people might think. In fact, I’ll go out on a limb and claim it can actually work for most classes in some form. And that is the final caveat. There are almost an infinite number of variations and ways to implement specifications grading. This is why I will only be focusing on essentially one version to demonstrate the core principles. I ask that you focus on the principles and not the specifics!
With that out of the way, let’s define specifications grading. Specifications grading refers to a grading system where clear standards are established for each grade level. This means that a student who earns an A met specific standards that differ from the standards set for a B, C, and so on. While this sounds simple and obvious, when we consider most grading systems used across the university, it becomes clear that these systems explicitly establish standards for an A, but a B, C, and down is determined by how far the student was from an A. Though unintentional, we often end up grading students by what they have not achieved.
Defining standards by what is not achieved is most obvious in courses based on points. A correct answer will earn the student the maximum number of points, and less complete variations of that answer will earn the student fewer points. This results in many different paths to earning grades lower than an A, and the standards for a B, C, etc. are not really well-defined. I would argue that these challenges also exist for rubric-based grading, but they are often more subtle for those cases. Unsurprisingly, there are a number of interesting unintended consequences of this approach to setting standards. One of the most significant is that it’s not clear what a student who earns a grade lower than an A actually knows or can do. When a B, C, etc. are determined by how short a student falls and there are many different ways to fall short, it’s inevitable for both students and instructors to be confused about the meaning of these grades.
So what makes specifications grading different and how does this work in practice? I will briefly focus on three components that work together, and this is where it can get confusing. The first key element is to be very clear on your goals for the students in the course and the means by which students will demonstrate they have achieved a particular goal. This could be through traditional assignments, exams, self-reflections, quizzes, participation, etc.
Once you have the goals and methods students will use to demonstrate they have achieved the goals, you need to organize the goals into the various grade levels. Depending on the structure of your course, there are two basic ways to handle this organization. One way is to have each grade level correspond to doing a set number of the grading means (assignments, quizzes, etc). For example, if you assign six mini-writing assignments, students who complete two or three would get a C, students who do four or five would get a B, and students who do all six would get an A. The other way to handle the organization is by offering assignments of varying complexity and difficulty that correspond to different grade levels. For this option, students can essentially choose their own adventure and achieve each grade level by completing each unique assignment. When designing a specifications grading system into a course, it’s more common to blend these two options together.
At the granular level, an important design piece is that all coursework becomes pass/no pass. The grade for each assignment is based on whether the expectation is met (and the student passes) or is not met (and the student does not pass). This feeds naturally into the grade levels that instructors define for their courses. Achieving a particular grade level usually involves passing a set number of items or allowing multiple attempts at particular items to reach a passing level before the end of the quarter. This pass/no pass approach is key to scalability because it allows instructors to be strategic about where time is invested in student feedback versus having to do a lot of partial credit and rubric determination.
One thing to be aware of when designing the specific criteria/goals for each grade is to consider the path from one grade to the next. You should not create a situation where, if a student fails to achieve the requirements for an A, they automatically get an F. This is why I mentioned that your framework will probably combine the concepts of “doing more assignments earns a better grade” and “different assignments earn a better grade” so that it is hard to meet the requirements for an A without also meeting those for a B or a C on the way!
The last thing to emphasize is that both instructors and students have to be really comfortable with the idea that we use three levels of passing grades, A, B, and C, and they each have value and a role. There are very good reasons why a student might only want or need a B or C in a course, and establishing clear expectations to achieve these grades allows them to have better control over their time and effort. As much as we as faculty would like to believe our course is the most important thing in a student’s life and that they all should be striving for an A, this isn’t reasonable. As my father used to say, “It is important to know when good is good enough!” and a C is certainly intended to be good. More importantly, transparency on grade criteria and some level of control over their grade, time, and effort is critical for student success and something we should all strive for in any grading system. Specifications grading is just one of many ways to accomplish this!