This school year I had three preps (Algebra II, Accelerated Algebra II, and Pre Cal) while coaching soccer. My evenings were spent trying to keep track of the three different things I needed to discuss the next day as the one planning period I was afforded during the day was usually spent trying to stay on top of administrative issues.

This situation was manageable in the fall, though time consuming. So with soccer season looming in January I knew my tutoring time would be low to non-existant leaving me in dire need of a new system to keep my sanity while still being there for my strugglers.

So, winter break hits and by random chance I stumble upon Dan Meyer's website and found his testing system. Trained as an engineer, I applied a little scientific method and experimented with it on my Alg II students for the semester that has just ended, figuring Pre Cal could stand to be done the "traditional" way for a little while longer. I had some concerns.

  1. The kids wouldn't buy in
  2. I would cover less material than my colleagues
  3. Grades would be dreadfully low

Much to my relief, none of those concerns came true. The kids loved (as much as you can love testing), I covered the same if not more material than my colleagues with more challenging questions, and 0% of my kids failed the semester without any need for me to fudge the results.

For those of you who don't want to read all of Dan's material, here's a summary:

  • Test every couple of topics, don't let textbook chapters dictate your test dates. I tested roughly once a week.
  • Ask 2 or 4 questions per topic, every topic is worth 4 pts. Instead of one test grade, each topic is its own grade. A test over 5 topics counts as 5 entries in the gradebook.
  • After you cover 2 or 3 new topics, test 2 or 3 old topics on the same test. Harder questions this time.
  • The best grade of those two attempts counts, students who get 4 pts. both times get a 5 in the gradebook.
  • Make the students track their progress. Give out some trinket as a reward for double 4s. I used stickers. Kids love stickers.
  • Give two official attempts at each topic. If a student still wants to improve their grade, they have to do so on their own time.
  • If a student elects to do that, grab an index card, scribble a couple problems related to that topic and see how they do. If they improve, delete the old grade. If they don't, they don't.

It took a while for the kids to get the hang of it, but once we got to the third test they were all in. A lot of them liked knowing where they were at grade wise. I threw out homework for the most part and took that percentage of their grade and made it participation. Part of which was keeping a notebook for their notes, handouts, problem sets, etc. They could keep the notebook in a bucket in the room, and it was amazing to see how many more paid attention when they had something to write on for a change. I could (and might) write a whole post about just this aspect.

So after tackling it for a semester, I noticed some things.

  1. It was still a lot of work. I had to rewrite my tests to match the system, which meant I'm cranking out a lot of these. This was my first year teaching Pre-Cal as well, so I'm having to generate those from scratch too.
  2. I could rid myself of spoon fed reviews and didn't have to waste whole class periods walking them through the test before I gave the test.
  3. Lots of kids bought in on the grade redemption. Many many weeks I was so happy to see kids turn 0s and 1s into 4s.
  4. Tutorials were very smooth. Rather than putting the work on me to find something to review, the topic checklist kept the kids very aware of what they should retake.
  5. Grading these tests was a dream. I could crank out test grades for my 56 Alg II students in less than an hour. Grading Pre Cal was insufferable in comparison.
  6. The running topic list made generating a final a 5-minute affair. I simply opened all my tests and grabbed the topics I wanted.
  7. There is potential to mislead kids on what their grade is. My school is on a six week system, and we're either issuing report cards or progress reports every three weeks. So in my gradebook I assign each topic 5 points, but if I don't have a chance to test a topic twice before a grade cut off, grades will be artificially low. If I leave it as out of 4 points, grades will be artificially high then the kids get upset when their grade "drops" later for "no reason" now that their 4s they couldn't make 5s are just 80s instead of 100s. At the end of the six weeks I would find myself needing to give extremely short 2 topic tests in order to cut things off nicely before grades were due.
  8. I developed a 3 + 3 = 3.5 rule, otherwise kids who were showing decent competency (and I think 3 out of 4 is competent) would still be considered failing.
  9. Partial credit is hard. It's difficult to reward good thinking but wrong answers without resorting to handing out 2.5s, 3.2s and so on. Perhaps a more ideal solution is a 10 pt. system, there's a little more head room to play with there.
  10. Bonus questions aren't compatible. In the past I liked to throw in bonuses for 2 pts. But 2 pts on a 100 pt scale is hard to award when you're limited to 5 pts.
  11. Oh the data! My gradebook auto computes an average for every assignment and displays it at the top. Extremely quickly I can tell what topics they find challenging and what topics they ace (and perhaps they should be challenged more at) and where I should retool in the future. Far more useful than knowing the average on the Ch. 4 test was 82%. Ch. 4 could be 8 topics for all I know.

Below are example tests, my final, pictures of student notebooks, and screenshots from my gradebook.

Test 6, Test 7, Test 8, Spring Final



AuthorJonathan Claydon