Problem Set D Researchers measured urinary beta thromboglob

Problem Set D Researchers measured urinary beta - thromboglobulin excretion in 12 diabetic patients and 12 normal control subjects, reporting their findings in Thrombosis and Haemostasis. The following measurements are Data Set 313 in A Handbook of Small Data Sets: 1. Do these measurements appear to be samples from symmetric distributions? Why or why not? 2. Both samples of positive real numbers appear to be drawn from distributions that are skewed to the right, i.e., the upper tail of the distribution is longer than the lower tail of the distribution. Often, such distributions can be symmetrized by applying a suitable data transformation. Two popular candidates are: (a) The natural logarithm: (b) The square root: Investigate the effect of each of these transformations on the above measurements. Do the transformed measurements appear to be samples from symmetric distributions? Which transformation do you prefer? 3. Do the transformed measurements appear to be samples from normal distributions? Why or why not? 4. The researchers claimed that diabetic patients have increased urinary beta - thromboglobulin excretion. Assuming that the transformed measurements are samples from normal distributions, how convincing do you find the evidence for their claim?

Solution