By Gabrielle Aizin ’19 and Maqadus Sakhi ’19
Just when students thought that they could take a break after a long year of statistics, they are in for a real surprise. Every year the AP Statistics teacher, Mr. John Cadwell, does an end of the year project to connect statistics to the real world.
“Students will soon notice that statistics helps them with many college majors and professions that they may choose in the future,” said Mr. Caldwell.
The purpose of the project is for students to conduct an experiment that relates to the real world and then use the skills they learned throughout the year to analyze the results. The project is the final requirement for all of the three AP Statistics classes. A major topic that was stressed in the class was significance tests. These tests allow students to determine if the probability of an event occurring is significant or not.
“Introductory statistics curriculum does not include any further topics than what was taught, so assigning this project will help students further their understanding of the topics covered and relate them to real life situations,” said Mr. Caldwell.
Students are required to choose a question to investigate, design an appropriate study or experiment, and then collect data. The data can be collected from a survey, observational study, experiment, or other sources such as the internet.
“I think it will actually be fun to conduct our own experiments. We spent a lot of time in classes learning about the do’s and dont’s of an experiment and now we are able to conduct one ourselves,” said Anastasia Krutikav ’19.
A study or experiment should follow all rules of a good design including being free of bias and randomization. Sampling bias is methods that tend to over- or underemphasize some characteristics of the overall population. There are several sampling biases that students need to be on the lookout for, including voluntary response bias, undercoverage bias, convenience bias, and response and non-response bias.
The challenging part of the project is to try to conduct a proper experiment with the least amount of bias. For example, when you are interviewing people that are of easy access it can produce convenience bias. Also, if you decide to let people fill out a survey online, this can produce non response bias or undercoverage bias because some people may not be able to access the survey which would alter the results of the experiment.
“My group and I are conducting our project on the amount of hours that a high school student procrastinates while doing homework. Almost every high school student procrastinates doing their homework, so we want to put these number into bigger perspective. The most difficulting thing we found so far is to limit the bias in our experiment,” said Maggie Chen ’19.
Once students collect the data they need to make inferences based on the data and summarize the data using appropriate graphical displays. The statistical analysis must have a confidence interval, a hypothesis test or a regression analysis. The final step is to present the research to the class in a five to eight minute presentation. The presentations will begin on May 31 and continue through June 8. Students are encouraged to use presentation enhancement such as videos, photos, audio and graphs.
“I really enjoy this project, it helps integrate and present all our knowledge of experiments and significances test in a fun way. The project also forced me to recall previous statistical terminology and methods,” said Noelia Cuautle ’18
Students are allowed to work individually or in groups of four at most and expand and improve earlier work or start something new. Students are advised take their time with the project and carefully plan every stage of the their project because each decision they make will have a major impact on the overall success.
“Since students have background knowledge on the topics taught, they can use this project to create a stronger mental model about the topics they learned this year and apply their knowledge to situations in the future,” said Mr. Caldwell.