Following are the data from Exercise 16.51 on bacteria counts, in millions, for different cases from each of five strains of cultured Staphylococcus aureus.

a. Use the Kruskal-Wallis test to decide, at the 5% significance level, whether the data provide sufficient evidence to conclude that a difference exists in mean bacteria counts among the five strains of Staphylococcus aureus.
b. The hypothesis test in part (a) was done in Exercise 16.51 by using the one-way ANOVA test. The assumption there is that bacteria counts in the five strains are normally distributed and have equal standard deviations. Presuming that to be true, why is performing a Kruskal-Wallis test to compare the means permissible? In this case, is use of the one-way ANOVA test or the Kruskal-Wallis test better? Explain your answers.

Strain A Strain B Strain C Strain D |Strain E 10 47 50 52 26 14 18 17 29 20 27 32 37 45 12 43 28 59 31 30 16