Statistics for Pathologists

This book provides a clear, accessible review of the use of statistics in pathology studies. It discusses exploratory data analyses, descriptive statistics, the use of comparative statistics, concordance analysis, categorical and continuous data regression analyses, count data, survival analyses and decision point and clustering analysis. Statistical analyses aid interpretation of research data in two ways: by providing descriptive measures and by making statistical inferences about population parameters and associations based on observations from a sample, a representative of the target population to maintain validity. Parametric tests are applicable to those data for which distribution assumptions can be made on the dependent variable of interest, the most common of which is an assumption of normal distribution. Hypothesis testing is the conceptual foundation of comparative statistics and statistical modeling. Pathology and laboratory medicine can run the gamut of sample size with some entities in anatomic pathology. Imputation is a powerful tool in many statistical approaches that allows us to put in values for missing variables based on the data. Sample size and power calculations are crucial components for both planning a study and reporting an actionable result. Sample size calculations are becoming an extremely common part of grant applications, which requires familiarity with them. More importantly, performing sample size calculations before collecting data and during the design of a study can insure that the results will be publishable regardless of effect size.