Letter to the editor: Statistical errors create bias
Hello, my name is Jeremy Meza and I am currently a student in AP Statistics under the tutelage of Professor Hager. I am responding to copy editor Dana Lin’s article titled, “We can embrace whoever we are.” Ms. Lin wrote, “Rather than taking on the impossible task of asking all the nearly 3,000 students at CVHS what they thought, I took a sample size of four students…” If Ms. Lin had taken AP Stat, she would have known that a sample size of four students does not accurately represent a population as big as 3,000 students.
Secondly, it is no surprise that the sample size Ms. Lin chose was not as random as she had hoped, seeing as she used a convenience sampling method, one of the most dangerous methods to use. This has resulted in selection bias, or undercoverage, as Ms. Lin has seen. In my professional opinion, I would use a stratified random sampling. If there existed a list of students by race, one could use a random number table to randomly choose an equal number of students from each race.
In conclusion, the results of Ms. Lin’s survey are biased and cannot be used to accurately represent our school. For more information on how to best perform a statistical observational study, please consult Professor Hager or Professor Eldridge or read Chapter 5 of The Practice of Statistic by Yates, Moore, and Starnes.
Jeremy Meza, senior
One thought on “Letter to the editor: Statistical errors create bias”
I am a student at Castro Valley High School and I am just glad to see articles/letters from students who value their education and factchecking as much as the people who are writing them. I commend the writer of the letter and the original publisher of the article for demonstrating a fabulous example of how journalism is supposed to work in real life. Kudos to both of you!
Comments are closed.