taking the plunge into standards-based grading

Posted on January 11, 2012 by Ian

So I’m committed: I’ve begun teaching Physics 291 (Intro Physics I w/Calculus) using a pure standards-based grading (SBG) approach. I still lay awake at night wondering what kind of train wreck this might be headed for, but it’s too late to turn back now. The fact that my enrollment is far higher than in past years for this course — full, at 60 students — doesn’t help. I still haven’t figured out quite how I’m going to handle reassessment…

Some initial thoughts about my experiences with and realizations about SBG:

Choice of specific standards is absolutely critical, and one key choice is “grain size”. I could identify a few larger, more general capacities to assess (extreme example: “I can use work and energy ideas to analyze situations and solve problems”). Alternatively, I could unpack those into a plethora of highly targeted standards (“I can draw velocity vs. time graphs for constant-acceleration problems based on a motion diagram”, “I can draw acceleration vs. time graphs for constant-acceleration problems based on a motion diagram”, “I can draw acceleration vs. time graphs for constant-acceleration problems based on a velocity vs. time graph”, etc. etc. etc.). Somewhere in between these extremes is a sweet spot that optimally balances specificity of feedback to the student with practicality of assessment and tracking.

I seem to be on track to have a bit over a hundred standards in this course, at a rate of about 6-8 per chapter. That’s 3-4 per class meeting, more or less. That seems like a lot, and more than many other SBG practitioners seem to have — but I’m having a great deal of difficulty combining them into more coarsely-grained standards without doing violence to my sense of what the “things” to be learned really are. To put it another way: The topics seem to naturally cleave along certain lines, and allowing that gets me to where I am.

Despite that last sentence, standards can be divided along various lines, and different ways of grouping sub-elements can align more or less well with the organization of my textbook and accompanying workbook, easier ways of assessing, etc. I initially brainstormed a list of standards, but have been doing some refactoring as I went through and correlated them with textbook sections and daily class plans.

SBG drives me to assess (and reassess) EVERYTHING I want students to seriously try to learn, rather than allowing me to sample a subset of the learning goals. I suppose I could simply not assess some of the standards and let them drop out of the grading scheme, but I currently feel that if it’s on the standards list, I ought to assess it. And that’s a lot! Which leads to my next realization:

Articulating learning standards makes me much more aware of what I’m actually asking students to learn (more than I would be with a traditional by-topics list), and there’s a freaking lot of stuff for intro physics students to learn. Wow. No wonder physics is hard!

If I want a relatively simple grade calculation — each student gets a 0-4 mastery rating on each standard, and the final grade calculations consists of averaging all those ratings and then mapping to a letter grade — then the number of standards per general topic had better be proportional to the topic’s importance, since that determines its weight in the overall grade. I find it tempting to split early chapters into many fine-grained standards (e.g., specific kinematics graphing skills, specific types of motion, etc.), but leave later chapters as more holistic standards (use the Impulse-Momentum principle to analyze collisions). Unfortunately, that overly weights the early stuff. I can either weight different standards differently, or unpack the later standards into finer-grained components… which is probably beneficial to both me and the students, but darn, it’s hard work!

Unless I want to box myself into having to assess each standard multiple times, in different ways (for different levels of mastery), or having different mastery scales for different standards, I’d better construct my standards such that only one assessment probe is necessary for each. That can mean peeling “advanced” mastery levels off of the top end of the mastery rubric and creating new standards specifically targeting those. For example: Instead of having the top mastery rating be reserved for “Can recognize need to apply this within a complex scenario and figure out how to connect to other principles” (which takes a different exam question than “Can apply to a straightforward situation when prompted”), I can have a separate standard for “Identify which principle(s) apply to a complex situation” and “Combine multiple principles to solve a problem”. Put another way: If every standard has an “above and beyond” level, I need to assess every student for that level of mastery on every standard, and that’s probably unrealistic. Better to have a few explicit “above and beyond” standards.

Reassessment is the heart of SBG — it’s what makes assessment formative, and lets students learn from their mistakes and keep making progress — but it’s also looking like the hardest part to implement, at least in my context (60 students, three 50-minute classes per week, the fact that giving up my free afternoons/days to a stream of reassessing students would kill my research efforts). I’ve been very cagey about not promising anything specific about reassessment yet in this course, but I can’t keep that up much longer.

The other big question, of course, is whether students really will do the work –reading, workbook, homework, etc. — without having those be graded. Most students do end up in the trap of running from deadline to deadline, only focusing on whatever is “due” next and prioritizing tasks by grade impact.

Stay tuned. This is very much a work in progress.

Posted in Learning & Teaching, Pedagogy, standards-based grading | 14 Comments

standards vs. authentic performance tasks?

Posted on November 9, 2011 by Ian

In my cogitations about whether and how to implement “standards-based grading” (SBG), I’m (still) wrestling with what appears to be a tension between (1) a focus on the factored, topical, individually assessable “standards” of typical SBG approaches, and (2) a focus on authentic, holistic, contextualized applications/projects/problems typical of things like “project-based learning” (PBL) and “problem-based learning” (also PBL). The former seems to require individual performance and accountability; the latter are often team-based and collaborative, providing yet another tension.

I find myself wondering about the feasibility of some kind of two-tier system, where (1) authentic, multifaceted, ill-structured PBL-type performance tasks are unpacked into (2) component/requisite “learning standards”; the learning standards are individually assessed, re-assessed, and hopefully mastered; and the overarching PBL-type performance task is then completed and assessed in its own way. Somehow, both levels would contribute to feedback and grading.

But, I worry about ending up with some Frakensteinian horror when the two are grafted together. “80% of the credit for ultimate standard mastery, 20% for one-time project grades” seems antithetical to SBG, and inconsistent. Building an additional “level of mastery” onto each granular standard to indicate “successfully used in a project” seems kludgy, and poorly aligned with the holistic nature of PBL.

Thoughts from experienced SBG implementers — or anyone else, for that matter? (Preferably not about SEO, though. Thanks. )

Posted in standards-based grading | 1 Comment

a first stab at “unit one” standards for Physics I

Posted on November 3, 2011 by Ian

I’ve been thinking extensively about “standards-based grading” (SBG) of late, ever since Andy Rundquist’s provocative dinner talk at the summer’s PERC banquet. (A summary of SBG and my general musings about it are fodder for a later blog post; for a taste, tune in to the #sbar Twitter hashtag, or check out the weblogs of Kelly O’Shea, Jason Buell, or the aforementioned Andy Rudquist.)

Last night I leafed through the first four chapters of Knight’s Physics for Scientists & Engineers — everything up to but not including forces — and scribbled down some potential “standards”, were I to be so bold as to try SBG next semester when I teach General Physics I w/Calculus. It’s definitely not a final set, but here’s the unedited list, presented for discussion.

Note 1: Some of these “standards” span or relate to more than one Knight chapter. I’m listing them here under the first such chapter.

Chapter 1: Concepts of Motion (i.e., “Basics”)

  1. Convert quantities between different units.
  2. Know and (where appropriate) employ SI units for all physical quantities used.
  3. Report and interpret numerical values for calculations or measurements, including appropriate units, unit prefixes, scientific notation, and significant figures.
  4. Determine values of kinematic variables corresponding to described, depicted, or observed motion, and interpret values by describing or depicting the resulting motion (including proper use of algebraic signs for direction).
  5. Produce, interpret, and interrelate graphs and motion diagrams of an object’s motion.
  6. Know and apply the definitions of fundamental kinematics quantities.
  7. Make reasonable order-of-magnitude (Fermi) estimates of physical quantities.
  8. Identify correct and incorrect expressions via dimensional analysis and/or limiting-case arguments.

Chapter 2: Kinematics in One Dimension

  1. Use the particle model and constant-acceleration kinematics formulae to produce a complete description of an object’s motion (numerical or symbolic) from partial information. [1D for chapter 2, 2D or 3D later.]
  2. Use basic calculus (derivatives and integrals) to interrelate functional forms for kinematic quantities.
  3. Use “free-fall” as a model to analyze real physical situations.
  4. Use the “inclined plane” as a model to analyze real physical situations.

Chapter 3: Vectors and Coordinate Systems

  1. Define and use a Cartesian coordinate system to describe an object’s location and motion.
  2. Interrelate the values of kinematic variables in two different coordinate systems (including translations, rotations, and Galilean relative motion), including “relative velocity” problems.
  3. Execute vector algebra (addition, subtraction, components, magnitude and direction) both graphically and algebraically.
  4. Represent, interpret, and interconvert between vector representations (graphical, component n-tuple, component unit-vector, magnitude & direction).
  5. Apply vectors and their properties where relevant when “using physics”. [Ick! But see note 3 below.]

Chapter 4: Kinematics in Two Dimensions

  1. Use “uniform circular motion” as a model to analyze real physical situations.
  2. Use “accelerated circular motion” as a model to analyze real physical situations.
  3. Use “projectile motion” as a model to analyze real physical situations. [See note 4 below.]
  4. Use angular kinematics in direct analogy to linear kinematics.

Note 2: All standards should carry the implicit rider “…and justify the applicability of the tools used, or identify when and why the task is not possible given those tools.”

Note 3: I’m unsure how to work in “applying” a specific thing (e.g., vectors and vector algebra) in addition to “knowing” it. I’m driving at the difference between active and passive vocabulary here: Yeah, so a student can do a vector algebra problem when presented with one, but will she identify the need to do vector algebra within an authentic context and apply it properly there? I could add a specific standard for that, but am afraid of proliferating standards.

Note 4: “Apply XXX as a model to analyze…” should be interpreted to include applying it to a piece or portion of a system or of an object’s motion, and stringing together multiple models or tools as necessary to solve a multi-part problem. (For example, inclined-plane as a model for a skier on a slope, followed by accelerated circular motion for a curved ramp at the bottom, followed by projectile motion for sailing through the air.) Or, should there be a separate standard for “Analyzing situations that require combining multiple ‘models’ or ‘kinds of motion’”?

I spent nine class days on these four chapters last time through the course (including an integrative pre-exam review day), so that comes to a hair more than two standards per day. Reasonable? Excessive? Thoughts about the grain-size of these standards?

Update: Since posting this, I’ve switched from Textile to Markdown for writing on this blog, since the best Textile plugin I could find for WordPress had some ugly bugs. Unfortunately, one of those bugs affected list numbering, with the result that the 21 standards above were numbered 1-21, rather than having the numbering reset for each chapter’s list. (The former may be preferable for this particular post, but the latter is technically correct.) So, numbers in the comments below may not correctly identify the standards they were meant to indicate. Apologies…

Posted in Learning & Teaching | 8 Comments

Is game-style learning fundamentally incompatible with school as we know it?

Posted on July 18, 2011 by Ian

My current scholarly “thing” is thinking about what we can learn about teaching, especially teaching physics, from the phenomenal power of video games to motivate, captivate, and teach. The impetus to ponder this comes from wishing that students would bring the kind of hard work, determination, creativity, resourcefulness, and collaboration to learning physics that they bring to playing, say, World of Warcraft. (For a blockbuster introduction to the topic, read James Paul Gee’s book What Video Games Have to Teach Us About Learning and Literacy.)

At the moment I’m more interested in lessons we can learn from video game design and take into more traditional, classroom-based instruction than I am in creating an actual video game that teaches physics. (The latter, however, is also a fascinating challenge to contemplate.)

In that vein, a definition by Bernard Suits (quoted in Jane McGonigal’s excellent book Reality is Broken) caught my attention:

Playing a game is the voluntary attempt to overcome unnecessary obstacles. (p.22)

The key word here is “voluntary”. McGonigal makes a case that if it isn’t voluntary, it isn’t a game, and many of the remarkable phenomena associated with game-playing disappear. The entire psychology changes.

Yes, attending university is in principle a voluntary choice, as is one’s major; but beyond that, we pretty much tell students what courses they must take and what they must do along the way to succeed, and keep them in line with grades and transcripts. Does that doom any attempt to make learning more deeply game-like?

What I’m getting at is that the very structure of our educational system frames learning activity as an externally-motivated, externally-directed, authority-laden series of tasks and assessments. I’m concerned that trying to embed a novel learning micro-environment — say, a gaming-inspired self-paced learning activity — into such a matrix could be doomed to failure, not because of the micro-environment’s worth but because of drastic dissonance with the matrix.

If I’m even more ambitious and try to construct an entire course as something analogous to a game, I still have to assign a grade at the end, and students know it.

Those of us who would like to experiment with gaming-inspired alternative paradigms and challenge some of our fundamental assumptions about what instruction should look like, and who don’t have the luxury of creating an entire parallel educational system to do our testing in, need to worry about such things.

Posted in Educational Research, Learning & Teaching | 2 Comments

playing a game

Posted on July 13, 2011 by Ian

“Playing a game is the voluntary attempt to overcome unnecessary obstacles.” — Bernard Suits, quoted in Jane McGonigal’s Reality is Broken

Is learning physics a game? Is doing physics a game? Does it depend on how obligated we feel to do any particular task? Is attending university voluntary (or compelled by social and/or economic considerations), and if so, does that make the whole endeavor a game? Taking any particular course may or may not be voluntary; doing homework, lab reports, etc. rarely is.

Why does this matter? Because in general, people like games, and often reach their best performance (think flow state) while playing games. Perhaps we ought to be learning from the game design industry.

Posted in Learning & Teaching, Physics Education Research | Leave a comment