I have been trying to participate in several twitter conversations today about grading, but can’t seem to satisfactorily get my ideas out in 140 characters. Hence blog post.
Have you read this article by Alfie Kohn on grading? If not, stop here and go read it. It’s a little long, but I’ll wait.
I recently was directed to it by Mike Skocko (@themaclab) because it’s one of the recent “quests” in his gamified media arts class (I’ve been poking around his game trying to get ideas to bring back to my class). The connection between the content of the article and the place I found it is important. I think that my biggest challenge to gamifying will be “what to do about grades?”. I’ll be posting a lot more about gamification in general, but for now here’s some ideas on grading:
First up, the academic side of things: I’ve already been doing SBG, and I like my system. The main component of my system that may be different from others is that I assess skills independently first, then together. But I think I want to change up my progress tracking. For each target, I want to keep the 0-4 scale, but change up the meaning of each number to represent the complexity level of the student’s mastery. Like this:
Level 0 = No evidence
Level 1 = Practiced the skill (self reported progress)
Level 2 = Basic assessment (you aced a multiple choice quiz)
Level 3 = Constructed response (you were able to explain your solution to a problem)
Level 4 = “In the Wild” Constructed response (you were able to apply this skill to a problem in which multiple skills were required)
Important: To get to one level, you have to complete the ones that come before it. One thing that is awesome about this is that it is one way to integrate a behavioral target with an academic one. To even get the first level, they are required to:
a) practice the skill (responsibility, independent work) AND
b) show me the practice work (organization, keeping track of practice work) AND
c) self assess whether they’ve made any progress (self regulation, my students are generally honest here, if they don’t get it they’ll say so).
Then, the second level is a multiple choice quiz. I keep up with these by doing them on the computer in a system designed by my colleague Stephen Collins (@socraticbrain) called Socratic Brain. I give a certain amount of quiz time each day and they use it to do whichever ones they are ready for (and by “ready” I mean they’ve already attained a level 1 in that target).
The third level is a constructed response, but just on that one skill. These can be “real world”, “fake world”, abstract, concrete, whatever. But the key is that it is one problem and just tests this one skill. The idea is that they can be graded quickly and the feedback is hyper focused. ”Kid, don’t worry about anything else besides the fact that in your third step here you used some faulty logic”. Here is an example of one:
To handle these, I have 5 copies of each version printed and keep them in a file bin. When a student is ready (they’ve attained level 2 credit), they request one from me, then work the problem on a separate sheet (that I also provide) and bring both papers back when they’re done. So far, this system has worked well. It can get a little hectic, but since they do one problem at a time, I can usually grade them and give feedback right then and there.
The fourth level is essential, but I have yet to master it. This is where the student gets a multi-skill task (initiated by me) and shows what he or she can really do. For instance, they may know how to graph a system, but can they do it if they have to write the inequalities first from a word problem? What about from a raw data set? Can they explain why a point is a solution in the context of a real world problem or only in the abstract?
The problem I’m having with this type of assessment is the “initiated by me” part. So far this year, this “level 4″ type of problem has only shown up on the quarter exam. But once I give feedback and the kid tries again and fixes mistakes, the report cards are already out. I still go back and change the grades, but it’s a hassle because I have to do an official grade change form in the office. So, I’d love to make this type of assessment a weekly thing, and I think I have a plan.
I will keep a spreadsheet with a list of these types of tasks and which skills they test. Each task will have a simple rubric, listing the skills with checkboxes next to each one. When I grade it, I check off the skills they clearly have mastered and give feedback. Then, when I go to enter the score in my gradebook, I just need to look at each skill and if they have level 3 and a checkmark for that target on the task, I award them level 4 status. Any other case gets you no improvement in the gradebook, but your reward is the feedback and the experience points.
I picture the gradebook looking like this:
This is basically how my gradebook looks now, but instead of a score out of 100 points (that gets averaged with all the others to form an overall grade), each learning target has a level shown in a progress bar. Notice the behavior target “Maintain an orderly binder” in there. It would of course need a separate rubric. Perhaps like this:
0 = you do not have a binder
1 = you’ve kept your binder neatly organized (handouts, notes, practice work)
2 = your binder has been consistently orderly (through at least one random check)
3 = your binder has been consistently orderly (through multiple random checks)
4 = your binder was orderly for every single random check this quarter (they’d have to earn this one back each quarter)
I think it should be fine to include this right alongside the academic targets. The only reason you might want it separate is to compute a final grade. But I don’t see any reason at this point to try to combine these scores into one letter at all. Instead, I’d much rather see a distinction made between “Ready to move on” and “Not yet” and that’s it. (thanks @JustinAion for the tip on the wording of those, sounds much better than pass/fail).
But in the event I have to stick a letter on it, I’ll need to ask what the purpose of the letter is. If it’s academic only, I’ll only consider academic targets (but really, the students will still get some effort credit as that’s what level one is all about). If it can include effort/behavioral skills, then I’ll consider all the targets when making my decision.
Any thoughts on this would be greatly appreciated!