Below are histograms and normal probability plots and Shapiro-Wilk test F-values of the Cortisol levels of the 31 cats in the study.

Shapiro - wilk p - value = 0.0001

Shapiro - wilk p - value = 0.6686
(a) Do these data meet the normality requirements for I lie two-sample t test?
(b) If we obtained larger samples, would we expect the resulting plots to appear more normal? Briefly explain.
(c) In a sentence or two, what would be the primary argument for using the Wilcoxon-Mann-Whitney test to compare Cortisol levels for trained and untrained cats?
(d) What is the primary drawback of using the Wilcoxon-Mann-Whitney test over the two-sample t test when you don*t "need" to use it?

Transcribed Image Text:

Trained 3.0 2.0- 1.0 0.0- -3 -20123 Normal scores Untrained 3.5 3.0 2.5- 2.0- 1.0 3 -3 -2 -1 023 Normal scores