Sneak Peak to Reduce Test Anxiety

Have you ever looked at the gender discrepancies on those who score a 5 on the AP Physics 1 exam? It’s nearly a 3-1 ratio! AP scrapped all of the data in 2021, so probably not. I started analyzing this data around 2019/2020 and was pretty shocked. It got way worse when you looked by race

“Surely not MY students” I thought. “I’m a female teacher AND I’m super aware of the issues around female performance in the physics classroom”

I checked my data. The same patterns persist.

So I dug a little deeper. I knew that I had female students who were on or above the playing field of some of my male students. What was going on that it was so hard for my female students to earn 5’s?

BY THE WAY… this is NOT a physics problem nor is it a math problem. Check out the gaps on AP Physics C and AP Calc…. WAY smaller

What I realized was it was their performance on the multiple choice.

Then 2020 offered an incredible opportunity. I could test my hypothesis by pulling the national data when AP had no multiple choice on the exam.

Guess what happened? The gaps were reduced.

In my college experience the classes I recall learning the most were the ones where exams were not “gotchas” but opportunities to deepen our understanding of the material. I had one teacher give legit take-home exams. It was nice, but not exactly a learning opportunity.

The next professor did something different. He gave us twice as many problems as would be on the exam a week ahead. We got together as a group and worked all of the problems over the week. The exam was “open annotated textbook” and the questions were ever so slightly different from the originals. The course was Physics 470 – subatomic physics. It’s the class I learned the most in.

The third professor who did something similar orally read us the exam the week before. He would leave out important details or specifics. “You have a circuit that looks like this… you will need to find the potential across two of the nodes” I also learned a lot in that course.

Taking all of these things into consideration, I’ve really modified the way I approach unit exams in my class. After seeing success with this model it has become a model I use starting on unit one and continue to implement until the semester mark.

I give students the entire test the day before the test.

Simple! I take off the part of the question that says “determine the _______”.

What do I mean by that? Here’s an example problem:

Now let me make this clear: students are expected to stow away all electronic devices before we begin so there’s no photos or google searches. Additionally, students are only allowed to use whiteboards. No paper. Nothing leaves the room.

Here’s another example

At first students are probably more stressed. The questions could be ANYTHING! The only thing students CAN do is EVERYTHING I want them to do! They have to draw force diagrams! Make graphs! Write out expressions for sum of torque and sum of forces. They have to consider all of the possibilities. And this is exactly what they need.

And the results?

First and foremost, implementing this strategy did NOT cause a huge increase in scores. In fact, some students still did kind of awful. What it DID do, what move up the mid-group from an anxiety score to a score that matched their classroom performance. This was particularly noticeable for.. you guessed it… my female students. The pretest in this case was a sample of questions from the unit one progress check. Students answered these in “test mode”. The post test was the actual kinematics exam. Below are the results:

Of course then I had to find out if it stuck. So when we got to the momentum test in January, I ran the data again. This time, there was no intervention: no sneak peek. And guess what happened?

What’s interesting is the number of male students who actually saw a decrease in score. I have a number of theories on this one, part of which being that they generally performed well on the first exam and so did not have that push to improve over time.

Due to the results of these data I have continued to implement this practice. After all, the goal is to get students working the problem, not searching for an answer!