Last year on a bit of a whim I decided to take some problem sets and turn them into poster projects. It worked well the first time and it was something I tried to do once a grading period in both of my preps. I liked it so much the tradition continued this year. But I've started to raise the demands. Usually they served as the single exposure to a task. So, say we're doing hyperbolas. We'll talk about the process, I'll graph a couple examples, maybe have them try one or something. Their assignment is a poster since hyperbolas are all nice and crazy looking. Each kid would do two graphs (I'd pair it with an ellipse or something). The side effect, however, is that exposure was pretty minimal. I would then assess them after only have them struggle with one, and typically if they were a struggler they got a neighbor to do most of it for them. Natural shocker, struggles really didn't know what was going on when it came for testing.
I modified two parts of my procedure to deal with this. First, I try to reserve posters as supplemental. So instead of doing a minimal-exposure activity within 10 minutes of learning something, it will usually follow after the skill has been assessed once and I've had time to digest and comment on their mistakes. It reduces mistakes on the more public poster. Second, I increased the workload on the students. Normally, a group of six would be given 12 problems and asked to graph 2 of them. Most recently in Algebra II, the whole set was required of everyone and I would change my phrasing to say "choose two that you will present." Basically, do everything, but two of them you'll have to recopy and do an additional task (paragraph writing, higher fidelty graph, etc). I think it's improving retention which is raising the competency I see on first assessments.
I also expanded beyond the poster concept and just made a giant wall of stuff. Now we have a Conic Wall (first seen here):
Rather than representing the total amount of a student's work on the subject, their work here is just a small sample. This strategy was also employed to improve our polar graphing exploration in Pre Cal. More on that soon.