Sunday, August 8, 2010

SBG Implementation: Power User Tips

School starts in a couple weeks and right now I'm wrestling with interim (benchmark) assessments. I'll let you know how that works out later. Until then, I've noticed a bunch of bloggers have been hashing out their standards-based grading plans.

Here are some quick tips that really helped me in the setup phase.

Topics and scales:

Cut breadth, not depth. At some point you'll find you have a ton of standards to teach. You will then realize that you can't teach that many standards. It is really tempting to try to lower your expectations so you can cover all your standards. Don't. Cut the content. Never cut depth.

Take a whole bunch of those standards and put them into the "I'm just gonna mention these" pile. When I say mention, I don't actually mean, just-say-it-and-move-on. You can spend the whole day (or more than that if you want) in your preferred method of instruction.

Usually, that means I tell my kids that they're going to need these for the state tests, but it's not going to be important for my class. I'll spend a day here or there loading some vocab, boring them with a Powerpoint, or doing an isolated lab and then just move on.1 You could certainly skip it entirely, but I'd like to give them a sporting chance at guessing on a 4-option multiple choice test.

Anchor your scale with the hardest assessments/expectations on your students. It's not uncommon for your students to need to take a common department final, a state-mandated end of course test, and an AP or AP-like test. Choose the hardest one, analyze the depth, and use that as your anchor. I didn't buy into this one until very recently, but I believe it now. It's a real problem if I'm setting my criteria based on my district benchmark, which is asking kids to read a passage and summarize what happened. Meanwhile their state test is asking them to make inferences.

Ask to see other teacher's tests in other districts. One of the things that keeps me up at night is the depth issue. I really worry that I am setting my level of expectation at a 5, while schools in Cupertino, Palo Alto, and Los Altos (insert your local high SES cities here) are asking their students to perform at a 10. I emailed about 20 teachers in other districts for copies of finals, benchmarks, whatever. Six emailed back. Since then I've seen three or four more. Mainly I learned that most teachers just use the exams provided by textbooks, but it did help me adjust a few topics and I also saw some really cool problems.

Start with the 3: I'm putting this here because MizT mentioned she didn't really get this until she read this book. I know your scoring system might be different, but you've got to start with the goal. Whatever it is you want your kids to learn, start there. Then work backwards to build the learning progression, which turns into your scale. Take a full or a half step forward to extend your scale. If it looks like backwards planning, or UbD, it's because it is. If you're going to teach that way, you should assess that way too right?

Your scales and rubrics are actually kind of useless by themselves. Sorry. I know you worked really hard on them. Your standards are meaningless until you define them with assessments and exemplars. There's a good example of that here, but it's gated. No matter how detailed and well thought out your scales are, you and your kids aren't going to really get them until they see some exemplars or they know how they'll be assessed. So don't sweat it if you don't have the wording perfect and you're not really sure if "Classify" or "Group" is a better verb. Spend less time working on your scales and more time working on the assessments.

Tests are for self-assessment.... I give tests. But I give them mainly for students to self-assess themselves so they can figure out their strengths—so they can replicate them—and their weaknesses—so they can work on them.

....and for you.
I need to have some info to play with to figure out what to teach next.

Most of your assessment will be invisible. You'll spend a lot of time asking questions as they're doing something, listening in to convos, or just peaking over shoulders. The more you need to interrupt the process, the less valid the assessment becomes.2 Your grades will rarely be attached to a specific "thing" which is why inputting grades by time, rather than assignment, is so useful.

Use your scales to help you give feedback. Leaving feedback was and still is one of my big weaknesses. I'm ok with written stuff but I've always been awful with oral feedback. My kids either do a "Great job" or need to "Work harder." Bleh. Your scales help. Leave feedback that specifically references the skills in your scales. "Looks like you're at a 2 right now, to move forward you're going to want to practice calculating density and using the correct SI units." And yes, you're going to want to teach them to be able to do this themselves.

Grades for the Whole Game, feedback for everything else. That's what the last post was about. I'm just reminding you. But think about it when you feel you need to grade every single thing. If it's not the whole game (which it usually isn't) feedback only.

Good luck new members of the SBG Borg.

1: More on mentioning from Grant Wiggins. This came via Twitter but I've lost the source.
2: I made that up as I was typing. I have no evidence for that statement.
3: If you're on twitter, follow the #sbarbook tag and jump in. On Monday Aug 9, they're starting How to Give Effective Feedback to Your Students. Most helpful book I've read in a long time.


  1. In terms of benchmark assessments, I had a new idea that we are probably going to try this year. Benchmark assessments will no longer have their own category in the gradebook. In fact, they won't even have a grade per se.

    All the individual skills will be scored out of 4 (from Far Below Basic to Proficient). Then, there will be two additional points for retention - at the 6-week and 12-week marks. When the students take the benchmark assessments, each skill that they still show mastery on will earn the retention point. Any skill that has been lost does not earn the point; however, they can remediate and retake the specific skills tests in order to earn missed retention points. This way, we can present the benchmark tests to students as an opportunity to show mastery (and earn corresponding points) of a lot of skills at once, instead of a high-stakes test that is entered as a single data point in their grade. I'd like to hear what you think about this idea.

  2. Hi Dan, nice to hear from you. Now for your question:

    I think I need some clarification. You already use a skill-based gradebook IIRC. So does that mean you'd have sections for normal "this is my class" skills and a separate "this is for the benchmark" skills? Seems strange so I'm pretty sure I'm not understanding what you've proposed.

    You've got my email so if you wanted to continue this there that'd be fine too.

  3. I do a lot of blogging on the use of clickers for formative assessment, mostly at the college level, but with some tips for K-12. As I'm still getting oriented to the K-12 world, I wonder how standards-based assessments might relate to the use of clickers? We generally recommend using them only to gauge students' understanding, by using them to spur peer discussion and ask students to articulate their reasoning. I know that some teachers use them to give quizzes, but I"m not as interested in that use, since it's just a technical implementation of the kinds of assessments we're already doing.

    So, in terms of using clickers for formative assessment to gain a window into student thinking, is there a way that you recommend using clickers to support standards-based assessment? Would the answer just be to use questions that are related to the standards and to the standards-based summative assessment measures that teachers have developed?

    Many thanks!

  4. @Stephanie,
    I use clickers, and am adopting SBG this year. I think it might be a useful tool to have a more real-time look at a student's understanding. How motivating would it be to show the student a graph of their improved understanding from (written) quiz to (written) quiz?

    I still haven't decided. I like the anonymous mode, but if it can be used as evidence of learning, I should probably track it

  5. @Stephanie
    Thanks for dropping by. I feel like Thomas Keller just stopped into my roadside diner.

    Warning: This is going to be the longest "I don't know" comment ever.
    "recommend using them only to gauge students' understanding, by using them to spur peer discussion and ask students to articulate their reasoning."

    That's the ticket right there. This is off topic but if I'm just using it to aggregate % on MC questions, I might as well let the kids stay home and take everything online. I could even be replaced by a no cost Google form that will send you to a different page based on your answer.

    As for the clicker question itself...I don't know. I've never used clickers as a teacher before so I don't feel qualified to answer this question. I'm not entirely sure how it'd be different from anything else that you'd do normally. I use index cards as low-tech clickers to help guide my instruction, making big changes, like running to the back room and pulling out a lab, or smaller ones, like going to Keynote slide 25, instead of 29. As you said, it's great for "Defend your answer." All these things are true though regardless of the grading system you use.

    I guess I'm stuck pondering how it would add value to a classroom that is standards-based vs. traditional. There are certainly some ease of data collection benefits, like attaching different questions to a single standard and tracking progress on that standard. You'd also be able to make adjustments as the teacher based on each standard a little quicker. But again, that's more of an ease of use thing.

    I would also use it as really quick beginning of class check for each student that will determine what they do the rest of the day. Come in, use the clickers for these 10 questions. Look at your results. If you're lowest on standard 1, you're going to..... If standard 2, then....

    Again though, I do that already, it just takes a little longer.

    I guess then the answer is I don't think it'd be different because formative assessment is formative assessment. There's an added value of attaching multiple questions/question types to a specific standard which would provide additional feedback and you could make adjustments based on the standards rather than specific questions.

    If you happen to stumble on back to this post, my weakness in answering this question stems from not knowing what clickers allow me to do now that wasn't possible before.

    Not just "makes things easier." I'm guessing you've got a blog post or two on that so a link would be great.