TALK: Power/Sample Size Calculations

Is it unethical to conduct studies that are too large or too small? If so, how do we find the right balance based on the available resources? This seminar provides an overview of sample size calculations and demonstrates how to use the free "PS" software to perform simple sample size calculations for the following study designs: - Comparator studies (two sample t-test, paired t-test) - Case-Control Studies (unmatched, matched) - Cohort studies (test of 2 proportions or odds ratios) - Survival Studies (log rank test) It will conclude by reviewing potential sources for estimates of effect size or variability and identifying situations where the best course of action is to consult a statistician.