Tess Abbey & Taylor Ligenza
Alcohol is a major problem for teens in the U.S. According to the National Institute on Drug Abuse, 72% of high school students admit having consumed alcohol during their high school career. How does Latin compare to these national trends? According to one of the posters hanging around our school, 66% of sophomores do not drink on a regular basis. So, after seeing these posters, we decided to find out whether the data from the Health and Wellness survey was correct. According to the survey, only 40.8% of Latin students admitted consuming alcohol. We felt that this percentage was low because of various flaws in the survey. For instance, one source of bias comes from the questions that asked students for their grade level, gender, height, and ethnicity. We believe that many students did not respond honestly to these questions because they felt that the school could easily identify them. Absurdly, the survey also asked students to give their waist size. We became suspicious about the survey results when the average waist size came back at over 5 feet!
Therefore, we decided to test the claim that only 40.8% of Latin school students have consumed alcohol. We believed it was an underestimation of the true proportion.
We created a survey asking students the following questions:
1) Have you consumed alcohol in the past year?
2) Have you consumed alcohol in the past month?
3) If yes to #2, how many alcoholic beverages have you consumed in the past month?
We began by randomly selecting 40 students to survey. Of those selected to participate in our survey, 27 responded and with those responses we were able to start our analysis.
After collecting the data, we found that 22 of the 27, or 81.5%, respondents answered yes to consuming alcohol in the past year and 17 of the 27, or 63.0%, answered yes to consuming alcohol in the past month. After carrying out some statistical tests, we found the chance of getting a difference as large as, or larger than the one we found in the first question of our survey was roughly 0. In other words, our findings are statistically significant and we have reasonable doubt to suspect the claims made by the Health Survey. In addition, we are 95% confident that the true proportion of Latin students who have consumed alcohol is 81.48% 14.65%.
Our analysis suggests that the claim made by the Health Survey is incorrect and that it underestimates the true proportion of students who consume alcohol. But these results reveal a deeper, more far-reaching problem than just the questionable claims the Health Survey has made. The results of our test should lead you to question how surveys are designed and conducted in general. They should make you question the statistics you see in everyday life and should encourage you to take a closer look at the statistics you may have previously been so quick to accept. The next time an ad uses a statistic to try to convince you to buy something, or a politician uses a statistic to get your vote, remember our analysis and take the information with a grain of salt.]]>