Schoolwide candy hunt teaches students about statistics


Photo by Christian Holt

Seniors Cameron Osbourne and Arunima Deogade measure Ms. Elder’s hand in the search for a candy prize. The competition is part of Mr. Baston’s statistics’ class.

Sydney Chin, Editor-in-chief

On the hunt for a candy reward, Mr. Baston’s AP Statistics students took their data collection outside the classroom in a schoolwide investigation to obtain a deeper understanding of how statistical models can be used to predict things in our everyday lives, and maybe even catch a thief.

Known for his hands-on and engaging approach to statistics, Baston has invented several ways to teach his students about the tangible, real-world applications of the subject. 

For example, earlier in the year, students competed in an annual competition named the “Z-games.” In the activity, students participated in a variety of mini-games like water bottle flips and mini basketball to gather data and learn how to use Z-scores to compare data measured on different scales.

However, Baston takes things a step further with a game unofficially dubbed “The Candy Thief Project.” Occurring since the second year Baston started teaching AP Statistics, students are given a list of teachers’ heights and room numbers along with a single handprint to be used to estimate and then identify a specific teacher.

Part of the fun for students is the in-class explanation of the game in which Baston mysteriously places three boxes in the front of the room supposedly filled with candy, only to then be dramatically ripped off the board in the surprise that another teacher had actually stolen the candy. 

“I think I get a little more theatrical each year,” admitted Baston. “I changed it up a little in the presentation of it this year, but the basic structure was not changed.”

As for his inspiration, Baston emphasized the importance of engagement for both students and teachers.

“We’re our own little islands,” expressed Baston. “I don’t really get to see much of the other teachers outside of building seven, even outside this hallway, so I thought it would be fun to get as many teachers involved as possible.”

Every year, an email is sent out to all faculty with the details of the game and all are welcome to join. This year, thirty teachers volunteered to be potential suspects for the game ranging from Ms. Marshall in 2-123 to Ms. Bubello in 7-210.

“Anytime you send out an email, there’s always some faculty member that’s chiming in to try and help you out,” noted Baston. “There’s a very collaborative spirit at this school.”

Out of this list of subjects, three thieves were chosen based on how unique their heights were. There was one culprit for each of the three statistics classes to make sure students didn’t give each other hints. 

“I made a list of the thieves and their heights and I did a one-variable analysis,” explained Baston. “I made a bar plot of that and the thieves that had the most unique heights. There’s a bunch of teachers that have the same height, so I made it so that there was the most difference between the thieves, just to maximize the chances of picking the right one.”

To win the game, students had to not only measure the length of the culprit’s hand but also their other classmates’ hand lengths and heights. This data was then used to create a linear regression model used to identify the thief based on the model’s predicted height. Students must also be able to calculate the relationship’s correlation coefficient, residual plot, and standard deviation of the residuals to get more clues to the actual thief. 

From the activity, Baston hopes students will gain a deeper understanding of the applications of linear regression models and how they work in real life. From setting the prices of products to determining the culprit of a crime, these models are constantly used to estimate outcomes. 

However, just as effective as these models are, they aren’t always perfect. 

“This activity was like the ultimate treasure hunt,” said senior Jeffrey Riche, a candy winner in period 4. “We learned that not everything is going to be on par with your estimate no matter how accurate your data is. For instance, with our class, the hand sizes we were looking for were 63′ and 70′, however the hand size of the thief was 62′. Not all data is going to be accurate.”

“That’s another very valuable lesson that I wish more people in the world understood,” added Baston. “Just because a model said something is so, does not mean that it will be so. It’s not perfect, but it’s useful.”

By the end of the week, the game was concluded and students were able to find out who the culprit was for their class if they hadn’t calculated it already. The thieves were assigned as followed: Ms. DeCosmo for period 1, Ms. Lezark for period 4, and Mr. Blaszczyk for period 6.

Several students were able, or close to able, to mathematically predict the thief from their data analysis. Period 1’s model ended up predicting that Ms. DeCosmo was the thief within an inch.

“I liked that the assignment combined our lesson with an interactive, enjoyable game,” said senior Afreen Mashadi, who was the second person to obtain the candy reward in period 1. “I was able to predict the correct height, but it took me a couple of tries to find the correct teacher since some of the teachers had similar heights.

In the future, Baston’s statistics classes will continue to prepare for the AP exam and participate in hands-on activities. Whether students are measuring the lengths of their hands or seeing how many water bottles they can flip in a minute, they can count on obtaining a new perspective of statistics and, of course, win some candy.