Data Set 9 in Appendix B contains measured levels of serum cotinine (in ng/mL) for two samples of nonsmokers: nonsmokers with no environmental tobacco smoke exposure at home or at work, and nonsmokers who are exposed to environmental tobacco smoke at home or at work.
a. Testing at the 0.02 level of significance, do the data support a claim that nonsmokers with exposure to environmental tobacco smoke have higher levels of serum cotinine than nonsmokers without such exposure?
b. Construct a 96% confidence interval for μ1 - μ2 where μ1 is the mean for exposed nonsmokers and μ2 is the mean for non exposed non smokers. Do the confidence interval limits contain 0? Assuming that raised levels of serum cotinine are undesirable, do the results suggest that exposure to environmental tobacco smoke poses a possible health hazard?