Most of my students in AP Physics 1 struggle with unit tests for a good part of the year. There are lots of reasons for this, including a lack of previous experience with physics, the difficulty of the material, and the difficulty of the tests. I do attempt to make each unit test model a mini version of the AP Exam. When a student does poorly on one of these exams, they often panic. Many are students who are used to making As in everything. They have usually done very well on tests. They don’t really have a recovery strategy when they bomb a test.  This year is no different, and perhaps a bit worse due to virtual/hybrid instruction and a reduced class time. I have a lot of students struggling with tests. 

To help out those students who bomb a test, I came up with “Second Chance Exams,” inspired by my previous experimentation during face-to-face instruction with “2-Stage Exams.” I learned about 2-stage exams from this article by Carl Weiman (Nobel Prize winner in Physics (and founder of the PhET Sims site) and others:

“Physics Exams that Promote Collaborative Learning”

The article was published in The Physics Teacher. If you teach physics and you haven’t joined the American Association of Physics Teachers, go ahead and join today. 

Here is my own summary of how I implemented the original approach. First you write your unit test, the “Stage 1 Exam.”. Mine are usually 10 multiple choice questions and then one or a couple of free-response questions. The source of the questions is usually old AP Physics B questions modified to be more like AP Physics 1 (make them more conceptual than calculation). I also use questions I wrote myself from scratch, and a few questions modified from other sources to be AP1-like material. The grades are scaled so that they approximate the distribution of points that corresponds to an AP Exam score. Even with the scaling, some students make a panic-inducing score.

I construct a “Stage 2 Exam” by taking some of the most-missed questions from the original test and rewriting them. If they were already conceptual questions, I change them into slightly different questions. If they were mathematical questions, I typically modify them to be more conceptual. In both cases, I focus on the mistakes in understanding that students typically make on the Stage 1 questions. 

Students take Stage 1, individually, in a typical manner (paper and pencil in my classroom, in a normal year). Then, immediately after they all finish (45 minutes to an hour into a 90 minute block), I pass out Stage 2, and students work together in their lab groups to complete that assignment as a group activity. Because of time limitations, I usually actually did Stage 2 at the next class period, although that is not the approach recommended in the original article above. 

This 2-stage approach works well for building student understanding, and improves grades. The Exam score is 75% Stage 1 (individual score), 25% Stage 2 (group score).  

In the current year, teaching Hybrid model, I came up with a different approach. Class time is greatly reduced. Many of my AP students are hybrid or full virtual and small-group collaboration is more difficult. I write and score the Stage 1 version of the exam in the same way, but I have been administering it online, using Canvas. For the second stage, I did something similar to what I describe above. I wrote a largely conceptual second stage as a Google Doc. I distributed it to the students who made a score of less than 80% on the first exam via Google Classroom. Students have to answer all of the questions to my satisfaction on the Google Doc to earn a replacement score of 80% on the Stage 1 test. Once they submit their work, I comment on what they wrote and return it to them. I only enter a grade if they complete the whole assignment to my satisfaction. Otherwise the Stage 1 Test Score remains the same. Students are allowed to work with a partner(s), look things up, or ask me questions. I do have lots of conferences (face-to-face and via Zoom) with students who submitted Stage 2, but still are stuck on some of the questions. Of course, some students ignore the whole process and just let their test scores remain low. Despite that, the system does seem to be working. It relieves a lot of the panic. I can focus students on their errors of understanding. They are encouraged to reach out directly to me when they don’t understand a second (or even a third) time. And the Stage 2 tests are relatively easy to grade. 

Here is part of a “Stage 1” Free Response question (modified from an old AP Physics B exam question):

And here is the “Stage 2” version: (I just noticed I did go ahead and ask the same question twice, probably because I saw so many calculation errors on what was supposed to be a review question)`

Here is an original Stage 1 question that was conceptual

And below is a Stage 2 version of the same question:

Here is a summary of my online version changes from Weiman Model:

My 2020 Online version is really a kind of modified test corrections. It seems to me to have advantages over test corrections:

1. The questions are new, so students can’t just ask somebody who got it right the first time. 

2. I can focus the questions on what the students didn’t understand, based on what they wrote on the first test.

3. The Stage 2 Tests are much easier and quicker to grade than test corrections. Everybody is answering the same questions. Everybody’s assignment is the same length. I can write an assignment of the length I want to grade, rather than grading everything that the student needed to correct. 

Let me know what you think! Stay safe!