In: Statistics and Probability
The Dean of a small college is investigating student preferences in class scheduling, and whether start times should be on the hour or the half-hour. She decides to survey a subgroup of the student body, consisting of 127 students (not everyone responded). Of this group 73 students supported starting on the hour, the remainder supported the half hour.
a. What is the 90% confidence interval on the proportion of students supporting classes on the hour?
b. Classes are currently on the half hour. Based on your analysis of the results, do you think the Dean should make a change?
c. If you repeated the analysis as a hypothesis test at 90% significance using a null hypothesis of p = 0.50, what would your result be?
Let p denotes the true proportion of students supporting classes on the hour.
a)
b) Since the upper bound of the confidence interval > 0.5, so at 90% level of confidence or at 10% level of significance, we can conclude that the true proportion of students supporting classes on the hour is significantly greater than 0.5, so there is sufficient evidence to suggest Dean to make a change.
c)
We can conclude that the true proportion of students supporting classes on the hour is significantly greater than 0.5, so there is sufficient evidence to suggest Dean to make a change.