I hate the new Blogger interface. It deleted my post that was supposed to be a draft and then published it so that only a title showed up. Boo. Luckily I can still revert to the old Blogger interface. If you clicked through to this before and saw nothing, my apologies.
After my
SBG brainstorm post, I really started to zero in on what kind of rubric I want to create. The main thing I want my rubric to do is to easily show anyone who looks at it (parent, student, teacher, administrator)
exactly what the student has done right and
exactly what the student has done wrong. I want my rubric to give clear feedback and possibly even cues or questions that instruct students on what to do next.
Is this even possible? I think the thing that has been holding me back is that I thought I could create a rubric that would work for everything (ala a writing rubric, a team work rubric, etc). But that is just not true. I'm looking for different skills for each concept I teach and so the rubric must be adjusted accordingly. Another hint that I must do backwards design: I have to clearly outline the outcomes I want students to achieve (a key concepts list), the method I will use to determine if they learned it (the performance event), and the method of communicating to students if they were successful (the rubric).
Another thought that has held me back is thinking that I should be able to look at a student's grade and immediately know where they went wrong
without looking at the rubric. I know, I know that makes no sense. But it kinda does. If you create a rubric
like I've seen elsewhere, such as
 4 Performs skill with no errors.
 3 Perform skill with minimal errors
 2 Performs skill with major errors
 1 Cannot perform skill
then technically, you could know how much a student messed up, but not at what exact point. So my new idea is actually better than the idea I was trying to live up to! Twisted.
I really liked
Lizzy's comment about giving students one point for different steps in the process like the AP world does on a regular basis,
Kristin's comment about giving students a basic, proficient, and advanced problem, and
Adam's comment about having students use rubrics to make test corrections.
I'm going to combine these ideas into a giant megamind idea!
My exemplar will continue to be on
system of inequalities and will now be accompanied with pictures.
Here we go:

#1 Basic 
#1 is the most basic problem. Students would be graded as follows:
 Graphing inequality 1 with correct slope and yintercept (2 points)
 Graphing inequality 2 with correct slope and yintercept (2 points)
 Determining if inequality 1 is solid or dotted (1 point)
 Determining if inequality 2 is solid or dotted (1 point)
 Plugging in test point for inequality 1 (1 point)
 Plugging in test point for inequality 2 (1 point)
 Shading in the correct direction for inequality 1 (1 point)
 Shading in the correct direction for inequality 2 (1 point)
 Stating an ordered pair from the solution set (1 point)
These are the most basic skills to solving systems of inequalities. #1 is worth a total of 11 points.
Moving on to problem #2:

#2 Proficient 
The only difference in this proficient problem is that now the inequalities are in standard form and must be put into slopeintercept form. Grading as follows:
 Solving inequality 1 with correct slope and yintercept (2 points)
 Solving inequality 2 with correct slope and yintercept (2 points)
#2 is worth a total of 4 points. Students will only be graded on the new part of the problem because they have already completed the basic skills on #1. Students won't know this and won't know that I'm only grading part of it. This prevents students from doing the bare minimum.
On to #3:

#3 Advanced 
The new skill in the advanced problem is that the student must write the system of inequalities from a word problem. Grading:
 Writing inequality 1 with correct slope, yintercept, and inequality sign (3 points)
 Writing inequality 2 with correct slope, yintercept, and inequality sign (3 points)
Students will then have to solve (like they already did on #2), then graph (like they already did on #1.) Problem #3 is worth a total of 6 points.
The entire performance event would be worth 21 points.
Students would receive a rubric at the beginning of the unit to monitor their progress along the way and to give feedback. They would check off skills as they learn them and write the date. I could use this rubric to grade quizzes along the way, using the bottom section for specific feedback. Then I would use the same rubric to grade their performance event at the end of the unit.
Rubric:

Rubric 
Does this seem clear and easy to understand?
Do the grade ranges seem fair? For example, a proficient student scoring 12 out of 21 points would receive a B on my rubric but normally a 12 out of 21 would be a 57%. That feels crazy. Am I totally missing the mark here?
I actually did grade this way, I just didn't give students a rubric. The assessment really was worth 21 points and I really did just give percentage grades like the aforementioned. The difference is that I would now be clearly communicating what I expect students to be able to do and clearly communicating how I am assessing those abilities. The grade is the only thing that is throwing me off. Am I out in left field here?