Friday, June 23, 2017

Advice for Getting Started with SBG

One of our teacher ed grads emailed one of my colleagues in search of SBG resources for a pre-algebra course he's developing for next year. He wrote:
When I had you for ____, you graded our assessments using a form of standards based grading. I remember receiving a paper back that listed what standards I had mastered and which ones I still needed to work on. I have wanted to try using a standards based grading system ever since I saw it in your classroom. Would you have any resources that you would be willing to share with me?
My colleague replied with several tips and resources, including Matt Townsley's growing list of scholarly articles SBG. He also Cc'd me, and while I was drafting my own response I realized it might be better to write it as a blog post. So here it is, for what it's worth: my list of suggestions and resources for getting started with SBG in the math classroom.



Hi ___,

I'm glad to hear you are thinking of making the move to SBG. I've been using it for some time now, and I'll never go back. Doug Reeves' comparison of traditional vs SBG principles (from Ahead of the Curve, p. 130) does a fair job of summarizing the reasons why:



 As you move to SBG, there are (at least) three important things to think about in advance:
  1. Which SBG Learning Targets (Objectives, Goals) will you use? 
  2. How will you manage the grades? (And what rubric will you use?)
  3. How will you manage reassessment?
Learning Targets: Early on, I had used lists of "I can” statements to organize the SBG assessments in my course, but I've moved away from that practice. There are a lot of reasons for that (see previous link), but one I've noticed recently is the fact that my students don't seem to use "I can" statements to work out what they're supposed to be learning anyway. Instead, they look at examples from assessment tasks I've given previously on a particular target (or those that were discussed in class) and say, "Oh, that's what I have to be able to."

So I've taken a simpler approach. I disaggregate my course content along two dimensions: content domains vs. skill areas.

For example, this summer I used SBG in my intermediate algebra course. I used eight content domains:

  00 = Functions
  01 = Linear Functions
  02 = Linear Systems, 
  03 = Exponential Functions
  04 = Logarithmic Functions
  05 = Polynomial Functions
  06 = Quadratic Functions
  07 = Rational & Radical Functions*
          *light coverage, so I lumped them together

and three skills areas:

  (a) Procedural fluency,
  (b) Concepts, Connections, and Representations, and
  (c) Modeling & Problem Solving.

to form a grid of 24 distinct target areas (e.g., Target 02(a): Linear Systems--Procedural Fluency).
Every graded assessment task was linked to one of those target areas. (Technically, "target" is not the right term, but it's the one I tend to use with students.) This has greatly simplified my life: I no longer have to agonize over the best way to cluster the dozens upon dozens of course objectives into a manageable-yet-meaningful set of assessment targets. Not everyone does it this way, but I've used it in several different settings now, and it works for me.

Managing Grades: You'll need to find yourself a gradebook option that lets you track reassessments and record scores by target instead of by assignment. Some find it easiest to do this with an old pencil & paper gradebook. There are some dedicated SBG gradebooks out there (e.g., JumpRope). Personally, I use a spreadsheet-based SBG gradebook that I've been developing bit-by-bit over the last few years. It's at a point now where it tracks reassessments and generates automatic grade reports from the grades entered to date, which I print out for each student after each major assessment. Here's a sample from this summer:



Because grades are recorded by target, not by assignment, I make sure each question on each assessment is clearly aligned with one or more learning targets, and I record separate rubric-scores for each target. To help students remember what sort of evidence is expected for each skill-area, I include a short statement to that effect under each Target label. When I grade the assessment, I evaluate the body of evidence the student has provided (i.e, their work on the tasks associated with a given target) and assign a single rubric-score for the relevant target(s) based on that evidence. I'll discuss my rubric next, but first here is an example from this summer showing how I set up my assessments.




Evaluation Rubric: As a rule of thumb, it's best to use a rubric with three or four levels. Anything beyond that and we are at risk of making fine-grained judgements that offer little in terms of increased accuracy and resolution. I use the following four-point rubric to evaluate students' work:
  P = Proficient,
  NP = Near Proficient,
  MP = Mixed Proficiency (or Making Progress), and
  B = Beginning.

I also use two other codes:
  I = Insufficient Evidence (I cannot determine the level of proficiency based on the work shown),
  X = Student was absent or did not complete the assessment.


Reassessment Options: Reassessment is an integral part of the SBG philosophy, but it's also one of the biggest challenges to implement.

In general, I have students complete written reassessment tasks, although I remind them that some reassessment is automatic as targets generally appear on several different assessment tasks over time.

Reassessment tasks, listed by target.
I prepare a set of parallel tasks for each target (not one per assessment) and save them as separate ready-to-print files labeled by both target and assessment (see screenclip). That way I can see at a glance whether I have an reassessment ready for a given target, and I can keep track of where the reassessment falls in the scope of the course. I print these on demand on tan or grey paper (to distinguish them from regular assessments), and only after a student has shown me they're ready. I grade the task and will go over it with a student afterwards if necessary, but I do not return reassessment tasks to students so I can keep reusing them.

There's a nice article (.pdf) by Rick Wormelli that discusses the reassessment in an SBG setting. It concludes with practical tips for managing "redos" in the classroom. I highly recommend it, and would add one more tip: Some reassessments may be done as oral interviews. If a student made a small mistake and can demonstrate understanding by convincingly explaining why it was incorrect, I will generally change their score without requiring a written reassessment.

I hope some of that helps as you work out the details on your own SBG implementation. I hope you'll share what you come up with!