In January 2011, I overhauled my grading system, and it’s still evolving.

In the old system, I tracked 10 assignments, 15 labs, and 4 tests.

In the new system, I still give 10 assignments and 15 labs (roughly).  I split each test (which would have been about 1.5 hours long) into three 30 minute quizzes.

The difference is that a student no longer gets a grade for the whole quiz. When I correct quizzes, I still mark each question.  I just don’t add them up at the end.  The student gets a complete/incomplete grade for each “skill” that was tested.

 

Why?

I do this because each question on the quiz targets a particular skill that I want the student to have.  I don’t need to know the average of their skills.  I need to know what to help them improve before I send them out to industry to pop breakers or start fires.  So I track each skill separately.  In the new gradebook, above, I can see right away who needs help with what.  Even better, the students can see for themselves what they need to work on.

How do you get a grade out of that? (aka “What are these skill ‘tiers’ or ‘ladders’ you talk about?”)

Each course has 4-6 units.  Each unit has 8-10 skills.  I hand out a skill-sheet at the beginning of each unit that lists them (see example below).  The skills are organized by complexity into levels 3 through 5.  Level 3 are skills that students absolutely must master in order to pass.  Level 4 are either less crucial or require combining skills.  To achieve level 5, a student has to propose their own learning, then demonstrate it.  The technical name for this is “conjunctive” grading, and I stole it from Kelly O’Shea.

If a student completes all the Level 3 skills, they have a score of 3/5 — in other words, 60% (our pass mark).  If they complete all the Level 3 AND all the Level 4 skills, their grade is 80%.

The catch is this: You can *complete* higher level skills anytime; but you can’t get *credit* for them until you’ve finished the previous levels. If a student has all of level 3, part of level 4, and all of level 5 skills under their belt, they get a 3/5 for that unit.  A student can’t get away with missing the fundamentals.  But the level 5 skills aren’t “wasted;” in the example above, when the student finishes the level 4 requirements, their score will jump to 5/5.   At the end of the semester, if all the Level 3 skills are complete, I average the unit scores and convert to a percentage.  But if there are outstanding Level 3 skills, what I type into the school’s database is “Incomplete.”  They can then negotiate with Student Services about a supplemental assessment or a learning contract.  Guess what’s on the supp?  You got it — the outstanding Level 3 skills.

What Constitutes a Skill?

These 10 skills make up the AC Measurement unit.  A skill is either complete or it isn’t.  There are no partial marks.  To get credit for a skill, it must be complete.

On the left are “theory” skills that will probably be assessed on a quiz.  On the right are practical skills that will probably be assessed in the shop. Level 3 skills are the requirements for passing.  Level 4 usually requires a more complex strategy or a combination of ideas.  Level 5 is not on the skill sheet.  Students who want to earn a 5/5 must propose a problem that they want to solve.  I’ll accept almost anything that isn’t identical to what we’ve done in class — the focus in on student decision-making, not complexity.  I sometimes put ideas for Level 5 investigation on the skill sheet (as on the example below) but the student must define the specific problem they want to solve.  Click through to expand.

Have I lost my mind?

If this sounds punitive and like it might be impossible to pass, I assure you that grades have gone up, and so has the retention rate.  The magic is in this one change: students can reevaluate any skill until the end of the semester.

Strangely, this makes much less work for me.

Before, I was tracking about 30 items.  Now that I’m assigning a grade to each skill, I have about 60 skills per course.  But I did a few things to simplify the system:

  • I stopped agonizing over partial marks, and gave every skill either a “complete” (1) or “incomplete” (0) grade

I don’t feel bad about not giving credit for a partially right answer, because it’s not the only chance the student gets.  When they’ve practised enough to feel confident that they’re ready, they ask for a reevaluation.  If they convince me that they have the skill, I update their grade to track their current level of understanding.  If not, they can try again next week.  This makes grading really speedy.

The corollary to that is that, in order to apply for reassessment, the students have to find and fix their own mistakes.  Remember, all I wrote on it was “complete” or “incomplete”!  I used to spend hours poring over their work with a fine-toothed comb.  Guess whose skills that improved?  Not the students’.  Now they use the fine-toothed comb themselves.  I grade less, they learn more.

  • I also stopped grading homework

I’m not convinced that you have that skill until I see you do it in front of me, without your buddies or a copy of a similar problem from the textbook.  You can have the grade when I’m convinced, beyond a shadow of a doubt, that you can do that thing independently.  That means either a quiz, or a demonstration in the shop, or a verbal presentation, or a screencast — I’ll accept any format, if I can tell that you did it yourself.  Homework might prove something to you, if you use it well, and it will definitely help you learn, or I wouldn’t bother assigning it.  But it doesn’t prove anything to me.

Why would students bother doing homework then?

I have other ways of making it worthwhile.  I often collect it and write detailed feedback, which most students want badly enough.  Also, if you don’t do it, you will probably not succeed on this week’s quiz.  Then you’ll have to apply for reassessment, which means submitting evidence of improvement.  Which means… you’ll end up doing the homework.  Students catch on pretty quickly.  But by itself, homework doesn’t prove anything, so it doesn’t get a grade.  And, let’s be honest, grading homework doesn’t make students do their homework either.  Lots of students will copy it or ignore it.  At least this way it eliminates the copying (there’ s no point, literally).  So the students who would typically copy the homework now simply don’t turn it in… which reduces their wasted time, as well as mine.

As mentioned above, they end up doing it later.  No problem — it’s no more than I would have graded before, except now it’s more honest.

How do you reevaluate students’ skills?

During shop period, they show me what they can do, and I chat with them about it.  If I’m convinced, I sign it off on their skill sheet.  Or they can submit evidence of mastery in some other format — I teach them how to make screencasts, phone videos, and other media for exactly this purpose.

Another option is to ask me to make them a fresh quiz.  When students select this option, they have to let me know by Tuesday at the latest by passing in their skills folder (sort of a portfolio) with evidence of what they did to increase their skill (practise problems, build a circuit, or any other demonstration that they have improved).   On Tuesday night I write back to them acknowledging the request, and asking for clarification or extra evidence if necessary. Wednesday, I make up quizzes with all the questions people have requested.  On Thursday afternoons, the students show up and choose buffet-style from the quizzes they need.  As students finish, I review their answers with them so they know what they got right and/or why they had trouble.

What about synthesis?

If I need the students to use two skills together, I list three skills on the skill sheet: one for each individual skill and one for the combination (as in the example above).  Also, level 5 skills are different from the others.  You reach that level by doing something that shows your understanding of the links between topics, and requires you to choose your own problem-solving strategy.  For example, last week we studied voltage doublers.  Three students teamed up and made a multistage circuit that multiplied the voltage by15, getting an output voltage over 100V.  I didn’t teach them how to do that, but I did teach them the skills they needed to figure it out.  It was up to them to take initiative, experiment, and decide how to check whether they had succeeded.

Is this the same as “outcomes-based” grading?

No.  Skills-based grading is a simple shift from an aggregate grade (“Chapter 2 Test”) to individual grades for each skill contained in Chapter 2.  It is agnostic about the authenticity or flexibility of the assessment. Although, in my case, the increase in data caused huge ripple effects in everything I do.  For details, see… the entire blog.

Does it work?

Looks like it, but new updates keep coming in.  Stay tuned.