The green bubble system has been described as “something of an artifact from the curriculum development tools we use here at Code.org. There’s certainly instances when they’re useful for tracking student progress, but the curriculum team wouldn’t advise just looking at bubble colors without ever examining student code more closely.”
Here is an FAQ answer on why all levels aren’t (or can’t be) fully validated.
So, there are a lot of different strategies teachers use. I tend to do what this teacher suggests (ie. check certain levels).
Some of my students caught on to that fact, so then I started telling them they had to do them all and I would check 1 or 2 levels per lesson, but I didn’t tell them which lessons I was going to check. That works for me.
This doesn’t meet everyone’s needs and code.org is aware of that. This post is one of their most recent attempts to help out, but if you have suggestions of other things they could try and implement, you can certainly contact them at support@code.org.
There are other discussion threads discussing teacher strategies for grading throughout the forum. Most of us have found a way to make it work for us, but also eagerly look forward to updated methods!
Mike