Every lesson this year started students out with an experience. After the experience, we helped formalize our students learning by layering definitions and formulas on top of the experience. One of the long-lasting effects of this instructional approach is that we always have a landmark activity to refer back to, often with a visual representation of the experience.
Activity: Gallery Walk
Get The Classroom Ready
Before students arrive for class, set up 6 - 12 stations around the room. Each station will have a reminder of a different learning activity that took place during the school year. The reminder might be a picture, poster, printed sheet, or a computer applet.
Which Activities Could be Included?
1. Can Joy Smell Parkinson's? (Inferential thinking)
2. Does Beyonce Write Her Own Lyrics? (sampling methods, bias)
3. How Much Do Fans Love Justin Timberlake? (stratified, cluster samples)
4. Does Caffeine Increase Pulse Rate? (experimental design, matched pairs)
5. Baseball Salaries (Central Limit Theorem)
6. Is Mrs. Gallas a Good Free Throw Shooter? (one sample inference for proportion)
7. Confidence Interval Applet (interpreting confidence level.
8. Is Yawning Contagious? (two sample inference for proportions)
9. Is One AP Exam Harder? (two sample inference for means)
10. Power applet (Type 1,2 error + power)
Gallery Walk
Put students into groups of 3 - 5.
Send each group to a different starting station.
Set timer (suggestion 4 minutes)
Students have a discussion about the activity at their station. When time is up, they move to the next station. They move until they have visited all the stations.
At the last station, allow each group to share out about that activity.
An Example
At the end of the gallery walk, we ask students to share out. Here is an exemplary response:
In the Parkinson's activity, we found out that Joy correctly identified Parkinson's status on 11 out of 12 t-shirts and we wondered whether this provided convincing evidence that Joy can smell Parkinson's or if this result could have happened purely by chance. We ran a simulation in which students smelled t-shirts and then randomly guessed the Parkinson's status. The dotplot gives the number of correct guesses for each simulation. The statistical content covered in this activity includes significance testing, null and alternative hypotheses, simulation, sampling distribution, p-value, Type 1 and 2 errors, etc......
How Does This Help Students on the AP Exam?
Imagine an AP Stats students who is in the middle of the AP Exam and encounters a free response question that is discussing a variety of sampling methods (refer to 2013 #2). To help answer part (a), they remember how in the Beyonce activity, almost all of the students overestimated the true parameter when they took a convenience sample. To help answer part (b), they remembered how we fixed the problem in the Beyonce activity by giving each word a number and taking a simple random sample. To help answer part (c), the student recalls from the Justin Timberlake activity how stratifying by row produced more precise estimates than stratifying by column.