The article "Increased Risk for Vitamin A Toxicity in Severe Hypertriglyceridemia" [Annals of Internal Medicine (1987) 105: 877- 879 (© American College of Physicians)] illustrates the importance of checking whether the appropriate conditions have been met prior to applying a statistical procedure. The data consist of the retinylester concentrations (mg/dl) of nine normal-individuals and nine Type V hyperlipoproteinemic subjects.

a. Assess whether the data sets support the condition that both population distributions have normal distributions with equal variances.
b. Test for a difference in the mean retinyl ester concentrations of the two groups using the pooled t test, separate-variance t test, and Wilcoxon rank sum test. Use μ = .01.
c. Based on your conclusions in part (a), which test statistic would you recommend to test for a difference in the mean retinyl ester concentrations of the two groups?

Type V Subjects 14 2.5 4.6 0.00.0 29 1.9 4.0 2.0 Nonnal Subjects 30.9 134.6 13.6 289 434.1 101.7 85.1 26.5 448