Question: Does the average number of hours of sleep a person gets decrease by taking more AP classes?
Hypothesis: Increasing the number of AP courses one takes will decrease the average amount of sleep one gets.
Explanatory variable: number of AP courses a person is currently enrolled in.
Response variable: number of hours of sleep the student gets on an average school night.
For my project, I surveyed students to see if there was a decrease in the average amount of sleep in students that are enrolled in more AP courses. My sample of students was chosen at random using the RandInt function on my calculator. I gave each student a number. Numbers were assigned alphabetically by first name. Then, I completed an SRS in order to find my sample of 25 students. I surveyed each student using a Google Form.
I surveyed students in a junior and senior AP Statistics class and a junior F.A.S.T class. My population was all juniors and seniors at Jefferson City High School.
This is not an appropriate model to use to make predictions. The scatterplot does not have a straight- enough condition. It is negative and nonlinear. The correlation coefficient is -0.138 so it is very weak. The residual is not scattered enough. Also, the variance is 0.019. Since this model does not satisfy the checklist, it is not appropriate to use to make predictions.
AP Courses Test
CI = (1.042, 1.998)
I am 95% confident that the true mean of AP courses taken lies between 1.042 and 1.998. I estimate the mean of AP courses taken to be about 1.52 with a margin of error of 0.478.
CI = (5.053, 6.307)
I am 95% confident that the true mean of sleep recieved lies between 5.053 and 6.307. I estimate the mean of sleep recieved to be about 5.68 hours with a margin of error of 0.627. Since our interval does not contain the value of 0, then it backs up our conclusion to reject the null hypothesis.
We reject the null hypothesis. We do have sufficient evidence to support the claim that the mean of sleep recieved is not equal to 7 hours. In samples of similar sizes, we would expect results more extreme than observed about 2 times out of 10,000 due to sheer random chance alone. This is statistically significant.