Overview: Methods that are easy on a pocket calculator and difficult on a software program are reviewed.You better understand what you‘re doing.Methodologies for which there is little software are reviewed.Pock.The first part of this title contained all statistical tests relevant to starting clinical investigations, and included tests for continuous and binary data, power, sample size, multiple testing, variability, confounding, interaction, and reliability. The current part 2 of this title reviews methods for handling missing data, manipulated data, multiple confounders, predictions beyond ob
|