Methods leading to Madness :)

**This was an email that I sent to one of the instructional coaches at my school for feedback and realized that it would be good here too - feel free to tell me what you really think in the comments :) **

Dear Instructional Coach,

I sent you a message a while back about how I would like to change my assessment practices. I am currently doing an online book study for Classroom Assessment and Grading that works by Marzano and next week we will start Formative Assessment and Standards Based Grading, also by Marzano. I have a pretty good understanding on the big ideas, but some of the detail issues still are evading my grasp :)

Ultimately, I believe an overall grade should reflect what a student has learned over the year, NOT how successful they were at accumulating points. Current practices are broken in that a student that understands the concepts but refuses to jump through the paperwork hoops "fails" the course and a student that has no clue but turns things in on time "passes" the course - that is just wrong in so many ways :( Also, I want my grading practices to be able to tell me exactly what a student does or does not know, broken down *by topic*. So there's where I am right now :)

I am pretty much contemplating two methods as of now... both are 100% assessment models and I would appreciate your thoughts if you have the time :)

**Note** In neither of these models does "homework" or "assignments" count in the grade. Students would still be expected to do warmups, homework, classwork, etc, but these are all ungraded categories. Ungraded can still mean that I take it up, look at it, write feedback/suggestions/comments on it, it just won't be assigned a numerical value in my gradebook.

Method #1 (Hybrid model)
Quizzes = 35%, Chapter/Unit Tests = 50%, Semester Final = 15%

Method #2 (More of a pure model)
Quizzes/Tests = 80%, Semester Final = 20%

In both methods, students would be provided with a list of learning targets for a chapter. Lessons would be taught as usual, although many opportunities for formative assessment would be provided through partner work, exit tickets, checkpoints (aka HW quizzes), but these would be ungraded opportunities. After a few lessons, a quiz would be given over the previous learning targets, with the grade broken down per target. Each learning target would have 3 levels of questions on the quiz (similar to the 3 levels of questions by Costa). Instead of one lump sum score, as in traditional grading, the quiz would have multiple scores, one for each learning target. This provides detailed information about which targets the students fully grasps, which ones are a work in progress, and which ones the student is essentially clueless on. The scores would be reported similar to:

4 = student successfully completes all levels of questions independently
3.5 = student successfully completes both level 1 and 2 questions independently and some level 3 questions with help
3 = student successfully completes both level 1 and 2 questions independently
2.5 = student successfully completes level 1 questions independently and some level 2 questions with help
2 = student successfully completes level 1 questions independently
1 = student successfully completes level 1 questions with help
0 = student does not successfully complete any of the questions, even with help

Now I haven't quite figured out how to translate those into my gradebook yet because our gradebook reports a pure percentage, so a 3/4 would convert to a 75%, even though a student scoring a 3/4 would be considered "proficient" and in my opinion deserves a high B, potentially even a low A. (Of course the discussion of "What does an A really mean" is a whole 'nother ball of wax)

So let's say a quiz in Alg2 has 4 learning targets (the equivalent to 3 lessons or so) and you earn the following scores:
- Solve absolute value equations....................score: 3/4
- Solve and graph 1-variable inequalties............score: 3.5/4
- Solve and graph compound inequalities.............score: 2/4
- Solve and graph absolute value inequalities.......score: 1/4

It's clear to me as a teacher that you have a pretty decent grasp on the first two learning targets, but you need to work on the last two. You would have the opportunity to get some help on those learning targets (go to the peer tutoring lab, come in for tutoring, do the assigned practice problems, work out of your workbook, go to a tutor, do *something* to show that you have put forth effort in relearning the material), and then show me the evidence of your learning and earn the chance for a "re-test" of the deficient concepts. This new re-test would then replace the previous score in the gradebook as it is the most recent snapshot of your learning progression.

At the end of the chapter/unit, I had thought about having a traditional test, graded in a traditional manner, but that's the main part I'm really unsure of right now. I see the value in having a test that you can't re-assess, but then that assumes that everyone learns at the same rate, which I don't think is valid. However, I'm not sure I'm comfortable with the idea they can continously reassess for 80% of the grade, nor am I sure of my ability to stay sane if allowing that much reassessment. This is really one of the last sticking points for me on deciding which model to pursue. In model #1, the test is treated traditionally (aka summative), but in model #2, the test is also reported based on learning targets, just a bigger "chunk" of learning targets at once and also allowed reassessment.

Anyway, if you've made it to here, thank you very much for reading! It helped me to get it "out on paper"

Sincerely,
Me :)