P05
The Effect of Stress on the Immune System: P-Value Comparisons
Paul Johnson, University of California san Francisco

Natural killer cells are considered to have an important role in immune resistance to tumor development. Natural killer cell data were collected from 112 subjects for a study carried out at the University of California San Francisco Hypothesis tests about the population mean of the subject's NK cells were carried out. P-values were calculated based upon the Central Limit Theorem, Type-2, the hybrid Bootstrap and the Bootstrap-t. A SAS(r) Macro was developed for 'B' Bootstrap replications and 'M' Monte Carl os to provide estimates of lower and upper 95% confidence bands for each of the four methods. The four methods result in p-value estimates that are limiting p-values. A Bayesian-type method is proposed, using all four methods, in evaluating whether or not a particular hypothesis should be rejected. The method provides for more confidence that a correct decision has been made. The macro requires base SAS and SAS/STAT software to run.