When I first started considering alternative grading, my focus was on how the approach would impact students. While that’s no longer a major motivating factor for me, I think a discussion on students is still important. It’s important to note that I consider students to be adults. Do they make bad decisions? Yes. Do I shield them from those bad decisions? To an extent.
Skipping (Parts of) Assignments
Taking responsibility for your actions, especially when the outcome is not what you were hoping for, is hard for pretty much everyone. There’s a whole specific cognitive bias, fundamental attribution error, about it where people will blame external forces for an outcome. I see this a lot in the courses I teach, partially because the way the K12 system is set up, if students fail, it’s the fault of the teacher. In college, that is no longer true (for the most part). How did this show up in specifications grading?
Focusing on the lab course, although similar things happened with the smaller writing course, managing multiple sections and TAs during the week meant that I depended on the Canvas course calendar to remind students of due dates. I did also post an announcement with a video showing students how to export the course calendar to their personal calendar AND change the reminder to more than 15 minutes before. This is important because all the submitted work had due dates, including token assignments and revisions.
Token assignments (prelabs, etc) and revisions were viewed as optional, which meant a good percentage of students didn’t do them. This became a problem at the end of the course when students realized they needed tokens to get to their desired grade, or at least the opportunity to get to their desired grade, and all the token assignment due dates had passed. I was already slammed with grading so I wasn’t willing to reopen them AND the assignments were no longer beneficial to student learning (they had 3 chances for each calculation, which was the only thing covered in prelabs).
I can’t lay the blame entirely at the students’ feet. While I did mention token assignments and had links to what the tokens could be used for, I should have asked the TAs to more explicitly discuss the benefits of tokens and the token assignments. I didn’t sell tokens as the life preserver they can be and so student blew it off. I teach a 1 credit course. I get it.
The part I didn’t get is when they would skip parts of assignments that were one of the criteria in the grade tracker. I separate calculations from short answers for several reasons, but one is so they can revise their short answers. However, in order to submit a revision, there had to be an original submission. Makes sense, right? Students would just…not answer them. Some they skipped because they were in a rush to get out of lab (paper link here) and then forgot to go back to. Others just didn’t want to do them. This became a problem when they needed at least 15 answered and they had only submitted answers for 9. Or worse, 14.
"I'm Only One Away! That's Not Fair!"
Determining course grades with specs can be tricky depending on your criteria. In my case, again, focusing on the lab class, the big issue was with the short answers. If a student had 8 high pass and they needed 9 in order to have a B+ vs a C+, it looks like it’s just one short answer holding them back. What isn’t seen is the opportunity to revise all experiment short answers for free a week after feedback is released. Or that for 2 tokens they can submit an additional revision. Or for 4 tokens they can convert one short answer from a low pass to a high pass. Or the optional lab that could replace a short answer.
I will say that having one criteria keep someone who would otherwise have a passing grade, end the course with an F didn’t sit well. If it was something like the short answers, it felt a little more self-inflicted but if someone panicked on the final. *cringe* With any grading system you have to be flexible for exceptions. I also give students a bail out option where they could choose to be graded using a traditional system. It was a one way decision however because I had to do the regrade of all submitted work. That was my boundary of labor.
Most of the emails fell in the category of one criteria resulting in a lower grade but it wasn’t only a problem for them.
DFW Rate
When I took over the course, I was proud of how much lower the DFW rate got. For those unfamiliar, that’s the percentage of students that receive Ds, Fs, or Ws (withdraw) in the course. I was averaging 1-3 students a semester with Fs. After the first pass with my original cut offs, I had 40 Fs which was equal to 20% of the class.