In 1969 researchers assessed the impact of allergens on sput

In 1969, researchers assessed the impact of allergens on sputum histamine levels, measured in micrograms per gram of dry sputum. In a sample of 23 smokers, ten were allergic and thirteen were not. The measurements of histamine-all values are correct-from their sputum: Suppose you want to test the null hypothesis H_0 : The two groups of smokers have equal levels of histamine in their sputum. against the 2-sided alternative hypothesis H_1 : Smokers with allergies have different levels of histamine in their sputum. You still do not like R so you ask another of your graduate students to create preliminary data graphics and numeric summary statistics for these histamine data. He gives you this plot from R (the horizontal lines depict the medians for each group): Analyze these histamine data using a 2-sample t test!. In your analysis, assume that the critical significance level alpha = 0.05 and that the degrees of freedom v = 9 . Make clear whether you use the equal- or unequal-variance version of the test. Justify your choice. What do you conclude from the results of your statistical analysis? Your graduate student analyzed these data using bootstrap methods. He gives you this normal quantile plot of the bootstrapped differences in sample means from R: What do you conclude from this bootstrap analysis? Next, your graduate student analyzed these data using the Wilcoxon rank-sum test. He gives you this output from R: What do you conclude from the results of this statistical analysis? Is your conclusion consistent with your conclusion from the t test? For these histamine data, which 2-sample test is appropriate? Why?

Solution