STAT 100B HWI Solution Problem 1: Suppose in a population of voters, the proportion of those who support a candidate is p . Suppose we get a random sample of 1000 people, and within this sample, the number of people who support this candidate is 530. (1) Calculate the 95% conﬁdence interval for p . Calculate the margin of error. A: ˆ p = 530 / 1000 = . 53. The standard error of ˆ p is p ˆ p (1-ˆ p ) /n = p . 53 × (1-. 53) / 1000 = . 0158. The 95% conﬁdence interval is [ . 53-2 × . 0158 ,. 53 + 2 × . 0158] = [ . 498 ,. 562]. The margin of error is . 0158 × 2 = . 0316. (2) Interpret the conﬁdence level of 95% (so that your roommate can understand you). A: Suppose we repeatedly sample 1000 people, and calculate the interval, the interval will change from time to time. In the long run, 95% of times the interval covers the true value of p . Problem 2: Continue from Problem 1. Suppose we want to test the hypotheses: H0 : p = . 5 versus H 1 : p > . 5. (1) Calculate the p-value.
This is the end of the preview. Sign up
access the rest of the document.