Patients with advanced cancers of the stomach, bronchus, colon, ovary or breast were treated with ascorbate. The purpose of the study was to determine if patient survival differed with respect to the organ affected by the cancer.
A one-way ANOVA with Organ as the discrete factor and Survival as the dependent variable is appropriate for these data. A normal probability plot of the residuals shows that they are skewed to the right. We transform the data to alleviate this problem. A square root transformation appears to make the residuals more symmetric, so we use the square root of Survival (ÃSurvival) as our new dependent variable.
The ANOVA table for this analysis indicates that the means of ÃSurvival are significantly different at the 5% level of significance with a p-value of 0.0002. Figure 1 is a boxplot of ÃSurvival by Organ. We can see from this plot that breast cancer patients seem to survive much longer after receiving treatments of ascorbate than patients with cancer of the bronchus, colon, or stomach. A post-hoc test using Scheffe's method confirms the results of the boxplot.