It’s hard to believe, but right now we are about 2/3 of the way through our semester, and with each week that passes, I’m getting more experience and insight with the use of specifications grading. Several people have mentioned on Twitter that they are following these updates with interest — and I don’t mind being the guy who goes first and makes all the mistakes. So here is an update on how things are going on the specs grading front.
First, some observations:
- We have now had two timed assessment periods in each of the classes in which I am using specs grading, and so every student at this point has worked a CORE-M problem, failed, and tried again. My first observation is just that this cycle of try, fail, try again has an importance in higher education that I had previously failed to appreciate. In higher ed, we talk a lot of platitudes about “lifelong learning” without seeming to apprehend what real lifelong learning actually entails, which is: try, fail, try again. We say we want students to be lifelong learners and then set up a system for assessing their work which is really nothing like how life really works (unless you are in a profession where failure truly is not an option, such as being a demolitions expert or an EMT).
- The second observation is an epiphany I had when a student was trying a CORE-M problem for the second time. The student came up to the desk to turn in his work and said, “I can’t believe I didn’t get this right the first time.” This pleased me very much. That student tried, failed, and then paid attention to the feedback and got it right the second time. I believe that’s called “learning”.
- I have had zero complaints from students about specs grading, and they’ve had opportunities to complain both in person and through opinion surveys. If there is any discontent from a student, that student is holding on to it like a bitter seed, in which case they’d probably hate it if we reverted back to traditional grading too because the student simply hates life.
- On the other hand, I have been frustrated at times when students don’t get what we are doing. One student, for instance, on a recent feedback question said that he felt stressed in the class because “you have to get things exactly right in order to pass”, even though I’ve stressed multiple times that passing does not mean perfection but rather B+ level work. So some students are perhaps stressing over things simply because they didn’t listen. Which means I need to communicate better.
In addition to these observations, I’m beginning to see some things that I didn’t do exactly well and can/should change for next time:
- The specifications themselves, I think, are way too numerous. They are difficut to enforce because there are so many of them. And the general idea is that students should take the specs for their work and translate them into habits, which is very hard to do if the list is a mile long. Next time, my list of specs will be at most half the length of what I currently have, and I will enforce them religiously.
- Students need a lot more examples of Passing and Non-Passing work up front, when the semester starts, to eliminate guesswork. I would have liked to have made more exemplars up this time, but I only had three weeks to design my whole system. So this will be a priority in the summer, to get a body of work ready to distribute at the beginning of the semester that students can use.
- It would also have been good to post all the assignments for the class at the beginning of the semester. There’s really no reason I can’t do this, and just let students start an assignment whenever they feel ready. Most students won’t want to work an assignment until we’ve discussed the material in class, but others may have life experience or more advanced coursework that makes them ready earlier than others. Again I think I fell victim to a tight time frame this time, but this will be another priority for the summer.
- I feel like my Learning Modules for both classes are a little too much like regular homework. In specs grading the idea is that you can allow students to demonstrate knowledge in multiple ways, and the only thing that matters is whether they eventually demonstrate they have mastered the learning objectives. There’s more room for creativity and individual choice than I am giving to the students at this point.
- Finally, I have to admit that I might be leaning toward having students take some kind of cumulative final exam. I mentioned last time that if students are working and reworking problems that assess core learning objectives during the course on a more or less continuous basis, there is less need for a comprehensive final. I still think that’s true. However, I had the following experience in the class recently: At the last timed assessment period in two of the courses, several students showed up and forgot to write down what CC and CORE-M objectives they wanted to be assessed over. Consequently several of these students re-took problems for objectives they had previously passed, and I guess it didn’t dawn on them that they had done these already. But they didn’t pass those when they tried again. This is “try, pass, try again, fail” as opposed to try, fail, try again. So this was bothersome because now I am not sure if students really have mastered these objectives or whether they only mastered them for that one assessment. That experience makes me think that incorporating a timed final exam would be a good quality-control check as students exit the course.
That’s the state of affairs with the specs grading for now. Overall both the students and I are very happy with the system — even when students want to talk about grades, we are talking about what constitutes acceptable professional level work, rather than quibbling over points, and I think it’s created a much healthier climate for the course.