Click here to download the PDF version of the transcript.
Stan Skrabut: Well, thanks for taking time to hang out with me and I appreciate you listening to this podcast. As I always know, you could be doing other things but you’re hanging out with me. It certainly does mean a lot to me. This week, we’re going to be exploring specifications grading. This topic kind of came on the scene somewhere around 2014. It’s a lot like, contract grading but there are certainly some differences to it. In my mind, it is a form of competency-based education. So far what I’ve read about specifications grading, I’m already a fan.
For some, there is a lot of stress when it comes to grading student work. I honestly think, after everything that I saw, that specifications grading can help a lot with reducing the stress but also maintaining the rigor of the content that you’re providing. We want to make sure that students are learning the content that you are providing. I think specification grading can help with that. I also think that it can certainly help learning. We’re going to definitely take a look at that.
The first time that I heard about specification grading was while reading the book Small Teaching Online. I talked about this particular book, Small Teaching Online, in Episode 28. I encourage you to go check it out, but it definitely caught my eye. I knew that I needed to go learn more about it. That way I could share it with you. Specification grading seems to originate from Linda Nilson’s book Specifications Grading: Restoring Rigor, Motivating Students, and Saving Faculty Time. I’ve included affiliate link to our book in the show notes, if you are interested. I have not yet read the book but so far, the reviews are very positive. It’s definitely a book that I’m going to be checking out in the future.
I honestly don’t think specification grading is really that new. Matter of fact, before I heard about specifications grading in the book, Small Teaching Online, I believe I was actually subject to it for 20 years while serving in the Air Force. In order for us to attain or maintain our qualifications in the various duty positions that we held, we had to pass our qualification exams. These exams consisted of written, oral and practical evaluations. In order to remain qualified, we had to pass all three parts. At the end of it, you either received a go or a no go. A go meant that you were qualified and you can continue on doing your duties. A no go meant you could no longer do your duties and you had to be reevaluated.
For the written and oral exams, typically it was an 80% cut-off. If you got below 80%, that phase of the exam was a no go. For the practical vows, you had scripted scenarios that you had to work through. Each scenario had specific tasks as well as a cut-off. If you’ve missed too many tasks, then you received a no go. If you missed critical tasks, you received a no go. Based on all that, if you received a no go on any of those parts, you were not qualified. You had to basically pass all parts. The specifications was very detailed in what you had to complete in order to meet that scenario or requirement. That’s what I see.
Apparently in her book, and I picked this up from higher ed article that Linda Nilsen points out some issues with current grading. It’s difficult to a lot points. Do I give this 93 points or is it 92 points? What’s the difference in the one point? Students naturally get spun up on the points. If you’re talking the difference between an 89 and a 90, that can mean a letter grade and students definitely want the higher grade.
The other fact is employers don’t care about the grades. They just want to know if you can do it or not do it. Our accreditors, Middle States, we’re coming up for a Middle States evaluation, they just want proof of learning and that learning is taking place. Let’s dive right into what specification grading are. It can be used for assignments. It can be used for tests but basically it comes down to two parts. Either it’s satisfactory or it’s unsatisfactory. It’s either a go or a no go. Either pass or fail or as Jason Mittell, who is a professor of American studies and film and media culture at Middlebury College, he has renamed the unsatisfactory part, unsatisfactory or no go or fail to not done yet.
I actually like this because that’s kind of how I do my quest. Either you successfully complete the quest or you get to do it again. Because you haven’t necessarily got it right this time. You didn’t fail it, you just didn’t successfully complete it. With this type of grading, students either get the full points or they get no points. It’s really simple to grade. Either they met their requirement or they didn’t. My time in Air Force, the way we looked at it is either you were able to do your job unsupervised or not. If you couldn’t do it unsupervised, then we’re going to work hard to get you so you could. As Mittell pointed out, this type of grading is very similar to the real world. Either it meets the specifications or you still need to do more work.
Another thing that came out in the reading when I was doing this Robert Talbert and article. I’m putting all these articles in the show notes, so you can go check it out. Robert Talbert, he notes in his blog that there is no partial credit, that you’re not giving partial credit. It’s either right or it’s not. This is very similar to my time at the Air Force Academy prep school. Either the answer was all right to include the work that you provided to get to the answer or it was not right, it was wrong.
We’ll talk about how to kind of build this in. It’s probably like, what are you talking about? One other thing in a presentation called Specifications Grading: A Promising Alternative, Molly Baker and Amanda Eichman summarize that students receive credit based on the number of work requirements and the specific work requirements they complete at a satisfactory level by the given due dates. This is how it all comes together. Now, how to actually build this? Good place to start is to have a rubric. That you are going to build a very specific and detailed rubric that identifies each task and has two columns. Either meet standards or does not meet standards.
The specifications that you’re building into this are: you’re aiming for B level work or higher. Definitely, the cutoff is B level. You don’t want them to get into the C level. You want them B level or higher. It’s really up to you how much detail you provide. For example, if you were to require a five-page essay, here’s some of the tasks that you could have on that rubric and this is not all-inclusive. You can certainly have more, but just give you an idea.
The first task, it needs to be five full pages and no more than six. If you’re looking for a five page essay, you want them five full pages. You don’t want less than five because students are pretty good. They’ll give you a four and a quarter if they can get away with it, but you want a full five. Which means it may drag over into page six but you certainly don’t want more than six. That either they did that or they did it. That less than 10 grammatical and spelling errors. You may want less or you may accept more. One clear focus topic supported by at least five peer-reviewed journal articles. They can have more, they can have 20 if they want but if they are four, that was a no go.
As you’re looking at this paper and you are writing your feedback, you can very simply go, is it five or more pages? Yes. Did they have less than 10 grammatical errors? Yes. Did they have a clear focus topic? Yes. Is it supported by five peer-review articles? Yes. Then that’s a go. They get full points for that assignment. It’s up to you how you weigh that assignment.
If you were to require, say, for a presentation that they build a PowerPoint deck, this would maybe look different. Maybe say, it needs to be between 20 and 30 slides. Text on each slide is clear and easy for the audience to see. You would define that probably a little bit. You know that it’s not less than 24 points for example but it needs to be clear and easy for the audience to see. Fonts, colors, backgrounds are effective, consistent and appropriate to the topic in audience. I’m sure that you’ve provided guidance for that and you can do that. It’s free of spelling and grammatical errors. Information on slide reflects understanding and effective summarization. Information has not simply been copied and pasted from another source. Either they do this or they don’t. Boom, boom, boom and they build this all up.
I’m already using these type of things in my class. I have what I use called Quest. I’ve talked about this in previous episodes that this is kind of a variation of this specification grading. Why? Because I’m using gamification as part of my class and assign Quests. My students get to pick and choose which Quests that they want to do but each Quest is either go or no go. Whatever it is, it’s typically a very simple straightforward assignment. Either they completed satisfactory or I give them feedback and kick it back and have them redo it. No harm, no foul. Don’t matter to me. I want them to successfully learn how to do whatever it is that I’m asking them to do.
Students get as many attempts as they want until they successfully complete the Quest. I remember one young lady, who completed or had turned in the Quest four different times, still couldn’t get it right but at the end, she did. She was thrilled. I was thrilled but she learned a lot in the process. That was the ultimate goal that she learned how to do it. That you don’t want everything to be so simple that everything is completed correctly the first time. Then probably the assignments that you’re giving are too easy.
There has to be kind of a happy balance. Second chances, this is absolutely I think essential for the specification grading. The authors of various articles talked about different ways they’re doing it. One of the things that’s really important and I believe is learning takes place more often when something is not done correctly the first time, than if it is done correctly the first time. In Nilson’s article as she was talking about specifications, she provided the opportunity for the second chances. The way that she does it, she doesn’t have so many resubmits. Students can also game that system that they constantly do resubmits. That just creates more work for them, if they don’t want to attempt to get it right the first time.
One of the things that Nilson does is she uses a virtual token system. At the beginning of the term, she gives students two or three tokens that they can use throughout the term for a variety of things. Here are some ways that she accepts tokens back that a student can use it for revising a no go assignment. Somebody turns in something that’s unsatisfactory, they can turn in a token and have an opportunity to redo it. Basically at the beginning of a term, she’s given two to three chances for redoing assignments.
She also takes tokens if a student knows that they’re going to bust the deadline. Now, for me, my deadlines are all on the last day of course because I do a lot of online teaching. I know that folks have, this is not their only primary thing they’re doing in life. That they have other obligations and they’re trying to weave their education in. I don’t have necessarily hard deadlines unless I’m doing something where there’s peer review. Then I will probably have a deadline because otherwise if everybody turned things in at different times, we couldn’t really do the peer review. That’s really the only time.
Other than that, I don’t have hard deadlines, except for the last day of class. Then that’s it, no more after that. I don’t take work after deadlines or after that last deadline, but that’s just me. Another reason that a student can redeem a token is to take up a makeup exam, that they wish to redo the exam. They turn into token, so they can attempt to get a higher score. You would have, as an instructor would have to build in a mechanism for tracking these different tokens. I use a Google spreadsheet for tracking all these things, these type of things. You can certainly put a link back into your learning management system. That’s only visible to you, so you can rapidly get to it.
That’s something that I particularly do but you can also weave this into your learning management system. Maybe make it an ungraded requirement, but just a place where you can kind of keep an eyeball on it. Now, Linda Nilson, she also has a mechanism for handing out tokens. What she does is, she rewards really good behavior. If someone’s turning in assignments that get graded to the satisfactory level and they do it ahead of the due dates, then she may also handout tokens. It’s up to you how you build that system.
Another one who wrote an article and one of the things that I gleaned from this article focus really on feedback. This is from a gentleman named Michael Kirkpatrick. His focus on feedback is really to help them with the content and support a growth mindset for guiding them to improve their product. That way, you’re not just writing it to justify why you’re taking away points?
So far the rubric or the way that you assess these assignments in specification grading, we talked about that, talked about second chances and feedback. There’s another piece that makes up this whole idea of specification grading. At least the way Linda Nilson has presented it. Since she happens to be kind of the central point on this and everybody else is making variations based on what she initially proposed. We’re just going to go with what she’s sharing.
What she also does is she creates assignment bundles. Now assignment bundles are stackable. You would create an assignment bundle at the very lowest levels that if you completed this set of assignments, you would get a D in the class. Then if you completed this additional set, you would get a C. If you completed this third set, you would get a B and this really fourth set you would get the A. If you completed all four sets, you got the A. You can group these different assignments and group these tests all together into an assignment bundle.
Ideally, everything is kind of related. You can look at this in many different ways. One is, you can create these assignment bundles around student learning objectives. The student learning objectives can be created in relation to Bloom’s Taxonomy, where at the very lowest level maybe you’re working on factual information, student recall and those things. At the very highest level, they are creating and synthesizing and analyzing and creating and really taking those high order skills in place. You would just create an increasing difficulty on the bundles throughout the course. The cool thing is the students get to decide what grade they want to work towards and work in that direction.
In order to get a specific grade, students would have to complete so many bundles. You may for example, have 10 bundles created. Like I said, each of these increases in difficulty, based on the content. You’re basically scaffolding content. It just keeps getting more complex as you go through. If a student added the 10 completed the first five bundles, they’ve got a D. If they completed bundles one through six, they get a C. If they completed one through eight, they’ll get a B. Then if they completed all 10 bundles, they would get the A. The work in there within the bundle has to be done to a satisfactory level.
That is where that specifications grading specifically comes into play, where if they need to pass an exam with 80%, that is a go. If they don’t, it’s a no go. Here’s a certain task, here’s the five steps, they complete all five, they get a go. Otherwise, they get a no go. The awarding of the bundle is they complete all the tasks with a satisfactory rate. As Talbert noted that specification grading allows students choice in agency and how and when they are assessed. It’s really up to you how you put this system together.
I use points. I don’t necessarily use bundles. I just require that students collect so many points and there’s cut-offs at each point. They started zero and they work their way up. Anything less than maybe, 600 points is not acceptable. Then anything after, it starts getting into the different grades. That’s how I particularly do it.
Some of the benefits for the students on this specifications grading, lots of different benefits. Talbert noted that it allows student’s choice in agency and how and when they are assessed, that this type of grading, they can walk out and show what they know. They specifically had to do something and they can better demonstrate that. There’s also being able to cope with failure or this idea of not being successful yet and how they deal with that? Their skills that they pick up. Their grades are based on actual concrete evidence of learning. They can attach meaning back to the grade because of this evidence that they have. The student is more in control of their grading. This is what they earned instead of what they think they deserve. Some things that make it very beneficial for students.
Now for instructors, there’s also some benefit. It saves time. You don’t have to like assess, “Well, they did five grammatical things and that’s going to be so many points.” No, it’s very simple. Do they have 10 or less? Yes. Okay, good to go, move on. It speeds up your grading in that aspect.
Now, naturally, you still have to provide the feedback but you can speed up the analysis of whether or not they were successful. As far as students go, it’s pretty straightforward. You can point out. It’s like, “Your paper was less than five full pages, it’s a no go. I asked for five, you need to put it between fully five or six and that’s good. The work, you’re doing this at the B level.” If they’re getting those successful completions on those various tasks, you can mark those off and you know confidently that you’re getting this at the B level.
It also reduces cheating and grade conflict. You don’t have to worry about those as much, depending on how you hand out your assignment. It also helps to map these back to the student outcomes a little more tightly. That is pretty much specifications grading. Like I said, I’m a fan because it’s very closely tied to competency-based instruction and I’m a fan of that. I like things to be very simple. Either you did it or you didn’t and if you didn’t, let’s do it again. I’m going to provide some feedback and help you learn how to do this better. Those are the reasons that I like. I’m moving in that direction more and more in the courses that I’m teaching and how I’m building out my assignments. Just a reminder, the way that Linden Nelson talked about specifications grading is for the grading itself, it’s either satisfactory or unsatisfactory. You can build this out in a rubric. Feedback is core, the ability to redo these is essential. Then also you have those assignment bundles where you can stack them towards a grade and students can pick and choose those type of things. That’s what we have for you right now. Before I let you go, here’s a plug for my book, Read to Succeed.