Saturday, May 29, 2010

SBG Implementation: Topic Scales

For those of you not active in the twitterverse, among the people I follow there's been an increasing interest in standards-based grading. Sam Shah called us an "inspiring ideological cult." I'm taking that as a compliment. 

With that in mind I'm going to explain my process for implementation.

Assumption: You already have a basic idea of what standards-based grading is. If not, go read every post by Shawn or Matt and come back.

Things to keep in mind:
  1. There is no canon for SBG. It's a still new decades-old idea if that makes sense.
  2. Context is everything. I've developed processes that fit my kids and style.
So here's my context: I teach 8th grade physical science in San Jose, California. That's intro physics, chemistry, and astronomy. Last year we had an API of 704 (800 is the goal in CA), with a similar schools rank of 4 and a statewide rank 3. 71% of my students are on free or reduced lunch and 44% are ELL.  We are 75% Hispanic. We border the East Hills which means we also have a group of kids that live in gated communities with swimming pools. I was originally the only SBGer at my school. This year there are two others (a 7th and 8th grade ELA teacher) and there are three more teachers (7th math, 6th math/sci, 7th ELA) that have adopted certain aspects of SBG.

What I've developed has been reasonably successful for my students, YMMV. Leave suggestions for improvement in the comments or send something through twitter or email.

Jargon alert: I'm going to use the word "standard" to refer to any unit of knowledge or skill you want to teach and assess. These are also variously referred to as learning goals or targets. A topic, also called a strand, is a collection of those standards that are grouped together in a meaningful way. Do not confuse standards-based grading with your state standards or standardization, although they probably will be related.

Step one in my adventure into standards-based grading: Building Topic Scales
Ok, maybe this isn't the first, first, first step, but it's the first meaningful one and most of us start here.

Even if you are not planning to implement standards-based grading, this is a valuable activity to do. You will become a better teacher just by going through this process.

For a quick review on why I chose to use topic scales, check out an older post.

What you're doing is setting levels of proficiency for each standard you hope to assess in your classroom. You are deciding on a target. My scoring system looks like this:
0 = No evidence of learning
1 = Can do most of the simple stuff with help
2 = Can do all of the simple stuff
3 = Can do all of the simple stuff and all of the complex stuff
4 = Can go beyond what was directly taught in class

I also have half points, like 2.5.

Group the standards into topics. This is a bit of vi vs. emacs. I'm not sure there's a right answer here.1 The math peeps generally opt for a straight skills list. I've noted my disagreement with that and am sticking to it. The downside of topics is that it is slightly more difficult for a student to tell what he/she needs to remediate. There's an extra step that some kids have trouble negotiating. On the plus side, topics facilitate connections between skills. Most of us have a good sense about what to group into a topic. In content classes like science or history, your existing units will fit pretty well. There are interesting ways to group science and history beyond Energy and Ancient Egypt but I won't get into that now.  If you're ELA you can have Persuasive Writing, Genre or the have a topic for each of the Writing Traits if you use those. Math might have Order of Operations, Percent and Decimals, Fractions, Probability. Each topic should have one or two really big ideas. I think somewhere between 10 and 20 topics per year is manageable. Too few and they start to lack coherence. Too many become unmanageable.

Start with the 3. What are the complex ideas/concepts/skills that I want my students to know or be able to do? Ideally, this is a process you're sharing with your department. Probably, you're flying solo here as I was so I drew from a variety of sources. I started with my CA standards, framework, and looked at released test questions. I've recently found performance level descriptors (pdf) that I used to adjust my standards a little. I also took a look at the standards for the high school science classes and used my own experiences with college and laboratory science to some degree. Be specific about what you want your kids to learn. "Learn fractions" isn't that helpful. "Add/subtract fractions with two-digit denominators" is much better.

Your primary concern here is twofold: What do my students need to learn? and At what depth do they need to learn it? If you're looking for help on depth, Bloom's and Marzano/Kendall's taxonomies were fairly helpful, as was Webb's Depth of Knowledge (pdf).

Backwards plan the 2. At this point you should have somewhere between ten and twenty topics. In my Forces in Fluids scale, my 3.0 included determining the density of an object, using Archimedes principle, and understanding how to manipulate the buoyancy of an object. You'll need to break these standards into a learning progression. Ask yourself, What do my students need to know to be able to do this stuff? Well they needed some basic vocab: density, buoyant force, pressure. They needed some measuring skills and a couple of SI units. They needed to understand that things sunk if they were more dense than the liquid they were immersed in.  That became my 2.0. You are trying to build a natural progression up your scale. Notice that the things in the 2.0 are needed to learn the more complex items. I'm not bombarding them with 50 different vocabulary words. For each topic I try to limit the vocabulary and memorization to the specific items that they'll need to understand the concept.

Set your 4. The 4.0 exists for one reason. If you directly taught it in class, it's memorization. It doesn't matter if it's a name, date, or the causes of the Civil War. Whether it's a simple idea or a complex idea, it's still our little birdies upchucking facts onto their papers. So your question for 4.0 is, Where do I want them to go with this? In my classroom, this usually takes a few different forms.  Sometimes I ask them to connect different information together in news ways, such as the relationship between metals having free valence electrons and also being conductive. I teach them the facts necessary to make that connection. I also try to teach them how to connect information.2 It's up to them to close the final gap. Other times I'll look at my high school standards and set the next natural step as the 4.0. In this case, I will spend a day or so directly teaching a few things. I violate my own don't-directly-teach-it rule here. Specifically, I wanted my kids to be exposed to balancing equations. You don't really need to know how to do it in 8th grade but our high school chem teachers spend forever on it. I didn't think it would be fair to require all students to learn how to balance equations so I compromised by setting it at the 4.0. I think of it as bonus knowledge. My other most common 4.0 style of question is just adding an extra half step onto an existing 3.0 level problem. The 8th grade standards for determining speed are very basic "You go 10 miles in two hours, what's your average speed?" problems. 4.0 might include finding an average of multiple trips or determining the speed you'll need to travel to obtain a certain average (I go 10 miles in two hours, if I want to maintain an average speed of 8 mph....). The higher levels on all of those taxonomies are helpful here as well.

Quick check: You have 10-20 topics. If you were to read the topic scales starting at the 2.0 and working your way up, it follows a natural and logical learning progression.  Your standards are grouped in a way that also makes sense. They're related and when students learn one standard in a topic, it helps them with the other standards at the same time.

Write out your assessments. I wrote out a pretty representative sample of all the questions I planned to put on my tests. It will be a back and forth process of writing problems based on your standards as well as revising your standards because you realize that's not quite what you wanted your students to do. I'm a huge-mungous believer that if you plan to assess it, you need to teach it. Look at your questions. If it states, "Analyze the effects of NAFTA on the economy and labor conditions in Mexico," then not only do you need to teach them about NAFTA, but you need to teach them how to analyze. Adjust your standards accordingly.

That's basically it. Look at your standards. Decide what proficient "looks like." Backwards plan how to get there. It took me a solid weekend to get a good rough draft. I used my first topic and immediately realized I had to revise things. Don't make 150 copies of all of your topic lists and laminate posters until you've got a few topics under your belt.

Science example: I have a topic called Atoms. The big idea is that all matter is made of atoms and that the existence of atoms can explain macroscopic phenomena. In the end, I want them to understand how temperature, pressure, and volume are related in a gas and the relationship between atomic motion, energy, and the state of matter. This is my 3. What do they need to know in order to get there? Well they've got to understand some basic vocabulary: matter, atom, solid, liquid, gas, pressure, temperature, volume. They've got to be able to differentiate between matter and non-matter. Since the 3.0 is centered around physical changes, they'll also need to know what properties of matter can and cannot be changed. They need to know the molecular motion of different states of matter. Those become my 2. For my 4, I ask them to explain certain phenomena that we haven't explicitly address in class. So we've talked about evaporation, but not condensation. We've learned previously why things float or sink, so I ask them to explain why a hot air balloon floats.

Non-science examples: Warning, my knowledge in these other areas kinda sucks so please don't flame me for getting some specific facts or terminology wrong.

Math example: For Kate Nowak's regression unit, Modeling or Regression would be a natural topic. The big idea might be that data can be modeled and then extrapolated using mathematical formulas. Kate might want her kids to be able to look at a graph and write a formula modeling the data. This is her 3.0. Going backwards from there, what do the kids need to be able to know in order to do that? Well it looks like they'll need to be able to identify different types of models. They need to be able to set up a graph. They need to be able to qualitatively explain what's going in the graph. They need to interpret different variables and constants. What would her 4.0 be? In our school we're asked to write across content areas so I naturally default to writing at this point. Perhaps she can give them a messier or incomplete data set and have them model it and justify why they chose that specific method. She may ask her students to create general rules for when to use each regression. She could ask them to identify specific examples of say, exponential growth, that weren't taught in class. Then gather some data and model it. 

Social Studies example: My big idea might be how the US government maintains a system of checks and balances using three branches of government.3 I want them to understand the historical roots for this system and how each system has served to enact change in their own way. What do my students need learn? They'd need to learn the basic functions of the three branches. They'd need to learn a few historical examples. 4.0? Students might be asked to contrast our system with a bicameral system. For higher level students you might use one of your historical examples and have them imagine how it might have played out differently with a different system of government.

Sticking points: The stuff that you'll struggle with.

You have a giant steaming pile of standards. You're going to end up with a lot of stuff. You MUST narrow the curriculum. If you are morally against omitting or skimming certain standards, leave a comment and I'll try to convince you. I'm going to assume we all understand the necessity. I usually don't omit standards entirely. However, I definitely underteach some. We have a fair amount of test score pressure on us. Instead of deleting, a few standards have been relegated to mere vocabulary and one or two just get mentioned and not assessed. There will always be a few one-offs because of testing but try to minimize them as much as possible.

Grain size. This is probably the big one for most of us. It's a struggle between being specific enough that you and the student can make a decision based on looking at assessment results and the simple management aspects of having to assess and track dozens of standards. This has been a constant tweaking process. So far, the most helpful thing has been analyzing student errors. Problem areas and hard to break misconceptions should be broken down a little bit more. When solving the mathematical problems, like rate, students would have trouble with using the correct units. I separated that out into its own standard. On the other hand, I lumped the characteristics of the planets into a single standard. Students came in with a lot of background knowledge so I didn't feel the need to have a separate standard for each planet.

The 4.0. Getting it "just right" ends up being tough. For some topics, it comes naturally to me. Others I struggle with.This is directly related to my content knowledge of that particular topic. My most common problem is getting overly excited with something cool and just blowing my students away. This isn't the time to have your students try to explain the double-slit experiment. History teachers really like counterfactual thinking questions. Those are super hard and you really have to be willing to devote a massive amount of time to teaching students how to do that. When you think 4.0, think half-a-step-up rather than a full level. This is definitely a context-dependent decision. You've got to know your kids and know what you teach and value.

Over reliance on verbs. The taxonomies are definitely a good guide, however simply adding in a "justify" or "predict" doesn't guarantee deeper conceptual understanding. Focus in on the big ideas and work backwards, don't try to build up the pyramid. I've made that mistake before.

Pro Tip 1: Don't go chronologically. Start with the topic you feel you've got down pat. It's really hard to create a progression when you only have a surface grasp of a certain topic. The physics and chemistry portions were pretty clear for me. The astronomy stuff I'm completely unhappy with even now.

Pro Tip 2: Be explicit with your performance standards. I've got students that can calculate the speed of an object on a test. I've got students that can figure out how fast they can skateboard. I want them to be able to do both. Write it out as two different standards and don't assume that just because they can do one, they can do the other. They can't and you won't know that unless you assess it directly.

Pro Tip 3: These scales aren't just for your tests/grades. They should guide everything you do in class. You should be able to directly link everything you do in class to something on your topic scales.

I hope this has been helpful. I'm working from memory here so I reserve the right to edit this page at anytime. If something is unclear, ask. If you've got a better method or something to add, I'd love to hear it.

Last second add: Scooped! Kate Nowak just blogged her checklist.

1: Unlike with vi vs. emacs. In that case, the correct answer is vi. I would like emacs if only it had a decent word processor. Ha!
2: I teach it pretty poorly, mainly with concept maps. If you've got a good strategy for helping students link disparate information together, let me know.
3: Wow. I'm really struggling with social studies. I guess I need to watch more History Channel.

Saturday, May 15, 2010


Shawn had an interesting post he titled Can O'Worm: Eat It, Salary Schedule. I disagree with nearly everything he posted (scroll down to the comments), but I appreciate his willingness to go out on a limb and say something that might not be entirely popular.

In that spirit, here's my can o' worms:

Teacher autonomy is vastly overrated.

Before you go crazy, I will say that we should have more autonomy in many areas. Our schedule, our textbooks, our budget, our ability to choose between equally good methods of teaching something.1

Here are three areas where we should NOT have autonomy.
1. Grades. You knew this was coming, seeing as how I'm assessment obsessed. It is 100% unequivocally NOT alright for an A in my class to mean something different than the teacher next door. Creating a clear definition of what a grade truly means is the essence of standards-based grading.

2. Pacing. Every same subject teacher should be teaching the same things at the same time. I'm not saying every day is exactly the same in every class, but on a topic/unit level, we should all be on the same pace. From a pedagogical standpoint, it makes collaboration and data analysis nearly impossible if I'm teaching about ancient Egypt while my neighbor is teaching about Greece. More importantly, there's a culture issue. Significant learning occurs...wait for it...outside of our classrooms. Maybe, just maybe, my students learn more in the hallways than at their desks. Having the same pacing creates a shared culture. Our kids should be able to talk about what they're learning, even incidentally. Even if it's to complain how bad Mr. Buell is at teaching.

Here's a not-entirely-imagined conversation:
Student 1: Mr. Buell SUX at teaching. I don't know get anything in his class. He just stood there and kept talking and talking about protons. WTF is a proton?
Student 2: Ms. W is kinda cool. A proton is that positive thingie in the nucleus......

There are strong teachers and there are weak teachers. Sometimes, our kids need to learn from someone else. Even if your school is entirely populated with master teachers, your style of teaching isn't going to be the perfect fit for every kid.

At our school, we have AVID and various remedial support classes. From everything those teachers tell me, it makes a HUGE difference when all the teachers of a subject are on the same track. Those of you who teach those classes, I'm sure you've run into the juggling act that is trying to help four different sets of kids learn four different things because their teachers can't agree on what to teach first.

3. Discipline. I'm known as a soft teacher. I break a lot of rules and let a lot of rules slide. My basic rule is that if you're learning and not hurting someone else's learning, pretty much anything goes. I've always been a "I should decide what is acceptable in my class" kind of teacher. You know what? I'm wrong. If I'm a hormonal 14-year old boy, I shouldn't have to learn 6 different sets of rules. I shouldn't have to learn that in my first period a raised hand means I just need to be quiet while in second period it means it means I need to turn and face the teacher and put my hands on my desk. My third period teacher marks us tardy if we're not in class while my fourth period requires that I have a paper and pencil out on my desk by the time the bell rings. It's ridiculous. If your school (like mine) has a lot of discipline problems, take a look at the varying levels of classroom norms a student has to negotiate throughout the day. Like with pacing, shared discipline creates a shared culture.

For me, number 3 is the hardest but I think most teachers generally have a problem with 1 and 2. I hate our rules. I feel like I'm fighting a flawed system by flouting the rules. I'm also undermining my fellow teachers and setting up my students to fail.2

If you're an English teacher and you're looking for a theme here, it's the idea of building a shared culture. If what I'm doing violates that shared culture, I shouldn't be allowed to do it. On the other hand, I should be allowed a whole bunch of autonomy within those confines.

As always, let me know what I'm wrong about and let me know what I forgot.

1: Notice the qualifier "equally good." If there's a better way and then there's my way, I don't get to choose my way.
2: That doesn't mean I won't work to change the rules. But until I do, we should all be following them.

Sunday, May 9, 2010

My first assessment failure

...and certainly not my last.

Before I knew anything about standards-based grading, I only knew that the way that we graded stunk. So I developed my own system. It turns out there's a reason teachers who have been in the classroom for only two years usually don't develop their own systems. My system sucked.1 I wanted to share it though to let you know what I learned from it and why the new system I use is much better.
I called it the Learning Matrix. Here's what a sample sheet looked like:
Learning Matrix

I based it off of two principals that still show up in my work with SBG:
  1. Separate learning goals instead of one lump score.
  2. Start with the end in mind and create a path to get there.

I decided to base it off of Bloom's taxonomy. Why Bloom's? I don't really remember at this point, other than we're always getting it thrown at us and maybe I figured I should use it. I recall at that time I used Costa's questions to create my question bank so I'm not sure why I didn't use that. Perhaps it just had a whole bunch of levels which I thought was cool. I don't know.

It worked somewhat similar to the skill checklist that many math bloggers use. The students had a set of questions or tasks that were specific to one of the grid boxes. Once or twice a week students would pull out the matrix and look at their grids. They'd select a box and find the accompanying assessment and work on it. The lower levels would involve definitions while the higher levels might be involve the completion of a mini-lab. After completion they'd bring it up to me and I'd give them a quick oral assessment. If they could satisfy my grilling I'd stamp it and they'd color in the box. There was also a pocket chart with the grid on it. Students would move an index card with their name on it to whatever they were currently working on. The idea was that a student could look for someone working on a higher level for help.

So what went wrong?

Ultimately the problem with this method was its basis on Bloom's. Bloom's is a classification system, not a learning progression. A learning progression lays out the steps needed to get to a learning goal. The current system I use creates a learning progression. How do I know this? Because I designed it that way. I started with my goal (in this case, the 3.0 concepts) and worked my way backwards. Let's say my 3.0 goal is for students to be able to calculate the average speed of an object. What do I need to know in order to do this? Some basic vocabulary, an idea of what motion is, some SI units, and how to measure stuff. That's the 2.0.2

I was waylaid by two other big things.

Number 1: I didn't teach the verbs. This wasn't necessarily a problem with my system so much as it was a problem with my teaching. I have a number of assessment pet peeves. Grading things that should be practice. Extra credit for non-academic stuff. Rubrics that focus on product guidelines instead of understanding. But gun-to-my-head I think I might say that my number one assessment pet peeve is grading something you don't teach. How in the world were my kids supposed to be able to "evaluate" when I didn't teach them how? I expected their mastery at one level to automatically gift them with the knowledge of how to do the next level. Even worse, I'd be surprised that they always got so badly stuck after starting a new level.

Now? I directly teach any higher order thinking that I will assess. If you're expected to "Compare/contrast speed, velocity, and acceleration," I'm teaching my kids how to make a comparison matrix. If I expect my kids to evaluate something, I'm teaching them how to create decision making criteria. Now, when a compare or evaluation question shows up on a test, they've got a set of skills to draw on. We need to directly teach thinking skills along with content knowledge. I used to think they just didn't get the content. They got the content. The thinking skill just got in the way.

Number 2: False equality. Not all topics need to be learned to equal depth. For the rest of their science lives, do all students need to understand the atomic theory? You bet. How about Archimedes Principle? Umm...not as important. The state of California dictates what I need to teach my students each year. That doesn't mean they all deserve equal time. In the new system, the atomic theory gets its own topic while Archimedes Principle falls under the topic Forces in Fluids. Some things (I'm looking at you polymers!) get demoted all the way down to just being vocabulary.

What's the take home message?
More than anything, that experience made me realize that this is going to be a long process. I felt like some sort of mad genius when I was designing the system. Two years into a formalized standards-based system, I'm still tweaking and learning. If you've been implementing a standards-based system you know what I'm talking about. If you're planning to, you'll know what to expect.

PS -  Don't forget to explicitly teach higher order thinking skills.

1: It still didn't suck as much as the way we regularly grade.
2: I could definitely be wrong in the setup of the learning progression. Perhaps learning doesn't occur in that order. However, my scales are fluid from year to year and I take solace in the fact that the research on learning progressions isn't all that solid.