ABSTRACT
Statistics has a long history in biomedicine, with interest growing in recent years given the increase in data availability. Desktop software has also evolved recently, and it is now possible to perform many common statistical tests quickly and easily. Measurements are an essential part of the evaluation. One type of variable can be more informative than others. Modern statistical desktop software has become easy enough to use and can therefore carry out the majority of routine statistical tests for health IT evaluation. There are two groups of statistical tests for hypothesis testing: parametric and non-parametric. Linear regression is a technique where only one continuous predictor variable is considered. Logistics regression is a regression technique used to analyze binary outcome variables. Subgroup analyses split data into subgroups and make comparisons between them; for example, in order to show that intervention has an effect on females over age 55, but not in other groups.
