I’m thinking about how to make assessments even lower stakes, especially quizzes. Currently, any quiz can be re-attempted at any point in the semester, with no penalty in marks. For a student who’s doing it for the second time, I require them to correct their quiz (if it was a quiz) and complete two practise problems, in order to apply for reassessment. (FYI, it can also be submitted in any alternate format that demonstrates mastery, in lieu of a quiz, but students rarely choose that option).
The upside of requiring practise problems is eliminating the brute-force approach where students just keep randomly trying quizzes thinking they will eventually show mastery (this doesn’t work, but it wastes a lot of time). It also introduces some self-assessment into the process. We practise how to write good-quality feedback, including trying to figure out what caused them to make the mistake.
The downside is that the workload in our program is really unreasonable (dear employers of electronics technicians, if you are reading this, most hard-working beginners cannot go from zero to meeting your standards in two years. Please contact me to discuss). So, students are really upset about having to do two practise problems. I try to sell it as “customized homework” — since I no longer assign homework practise problems, they are effectively exempting themselves from any part of the “homework” in areas where they have already demonstrated proficiency. The students don’t buy it though. They put huge pressure on themselves to get things right the first time, so they won’t have to do any practise. That, of course, sours our classroom culture and makes it harder for them to think well.
I’m considering a couple of options. One is, when they write a quiz, to ask them whether they are submitting it to be evaluated or just for feedback. Again, it promotes self-assessment: am I ready? Am I confident? Is this what mastery looks and feels like?
If they’re submitting for feedback, I won’t enter it into the gradebook, and they don’t have to submit practise problems when they try it next (but if they didn’t succeed that time, it would be back to practising).
Another option is simply to chuck the practise problem requirement. I could ask for a corrected quiz and good quality diagnostic feedback (written by themselves to themselves) instead. It would be a shame, the practise really does benefit them, but I’m wondering if it’s worth it.
All suggestions welcome!
Shifting Phases 
This is great, Mylene. I require students to use a credit system to sign up for reassessments that end up in the grade-book. So many hoard their credits for the end of the semester, however, that it doesn’t result in the sort of continuous feedback I want them to be getting. I love the idea of giving them the option of telling me which type of feedback they want, and not requiring anything for that situation.
I know I need to refine my system further, and this idea might just be what I’m looking for.
I like this idea. I use Canvas LMS for all my high school physics course quizzes, and use random selection of questions from question banks.
I’ll typically give the students two attempts with a small number of questions. Upside is they see lots of different questions, just not all at once. Downside is that one attempt might be easier than another, and give a false sense of understanding.
I can see giving an ungraded practice quiz first as a more powerful tool to get students to practice the work before scoring attempting a “for points” quiz.
I don’t do quizzes. Instead I have two assignments a week: prelab questions due on Monday that I grade Monday night and return at the start of lab on Tuesday, and design reports due on Friday, that I grade and return on Mondays.
The prelabs are a one-shot assignment—there is no possibility of redoing them or turning them in late. They also carry little or no weight in the grading, but are just there to get the students up to speed before lab starts. They seem to be working for that goal—the students are much more productive in lab this year, not trying to read the book and do the prelab exercises during lab.
The design reports can be redone, but I only give the students one week to turn in the redone report. I ended up with too big a stack at the end of the quarter last year when I didn’t have the deadline. Any error on the schematics triggers an automatic REDO grade (which turns to F if not redone).
Grading the design reports (which are about 5 pages a week for each pair of students) takes my entire weekend, and I’m about a week behind on grading redone reports (they are the lowest priority on my grading time).