STAtistics / Mr. Hansen

2/14/2011

Answer Key for Practice Test on #11.70, 11.74 (p. 637)

11.70	Let = true mean difference in no. of seeds detected (stratified − direct)
	H0: Ha:
	Assumptions for 1-sample t test
	SRS? (not stated, but must assume in order to continue test) normal pop. of differences? (not met; there is clear right skewness in the difference col., but n = 27 is fairly large, close to rule of thumb value of 30, so we proceed with caution)
	Conclusion: There is insufficient evidence (t = 1.34, df = 26, P > 0.19) to conclude that there is a difference in the true mean number of seeds detected using the two methods.
11.74	Let p1 = true proportion of males who are concerned about the possib. of AIDS p2 =      "         "          "  females            "
	H0: Ha:
	Assumptions for 2-prop. z test
	N1 and N2 both larger than 10 times sample size? (yes, if university is large, with at least 10,000 students—must assume this to continue) n1p1, n1q1, n2p2, n2q2 all > 5? (yes, all are in the dozens or higher) 2 indep. SRS’s? (“random” was given, so this assumption appears to be satisfied)
	Conclusion: There is strong evidence (z = 4.09, P < 0.0001) that the true proportion of female students at this university who are concerned about getting AIDS exceeds that for males.