Of the numerous available statistical methods, the list below presents a selection of procedures I am familiar with. It contains common and widely used methods (e.g., various tests or linear regression models), as well as advanced techniques (e.g., mixed-models) that I know in-depth. The grouping reflects the different data structures encountered in research.
Descriptive Statistics
Comprehensive tables of statistcs (location and spread of the data) for plausibility checks and reference
Tests
Hypothesis tests using parametric methods (e.g., Z test or t tests)
Hypothesis tests using nonparametric tests (e.g., Wilcoxon or Mann-Whitney test)
Permutation tests for exact p-values
Resampling methods
Analysis of Continuous Data
Correlation (Pearson, Spearman, Kendall)
Linear regression models, using “first aid transformations” (Tukey) and applying thorough model diagnostics
Group comparisons (ANOVA or Kruskal-Wallis test)
Nonlinear and robust methods
Nonparametric density or kernel estimation
Repeated Measurements
Analysis of Co-Variance (ANCOVA) models
Mixed-effects or hierarchical models (using maximum-likelihood, restricted ML, or robust methods)
Generalized Estimating Equations
Analysis of Categorical Data
Analysis of proportions/risks (chi-square or Fisher’s exact test)
Quantification of differences (risk difference, relative risk, odds ratio, number needed to treat)
Mantel-Haenszel methods
Logistic regression models
Exact-like logistic regression
Cumulative logit (or proportional-odds) models for ordinal categories
Analysis of Count Data
Analysis of counts or rates (counts per unit)
Quantification of differences (rate ratio)
Mantel-Haenszel methods
Poisson regression
Negativ-binomial regression
Survival or Failure or Reliability Analysis
Estimation of the survivor functions (Kaplan-Meier plot)
Log-rank (or Mantel-Haenszel or Mantel-Cox) and Gehan-Wilcoxon test of two survivor functions
Analysis of the hazard ratio (Mantel-Cox estimator)
Cox proportional hazards regression model
Weibull regression
Time-dependent covariates
Time-dependent coefficients
Competing risks and multistate models
Classification
Sensitivity and specificity of binary predictors
Predictive values and likelihood ratios
Analysis of ROC curves
Principal component analysis
Discriminant analysis
Trees and Forests
Reliability of Measurements and Raters
Measurements of reliability (Cohen’s kappa, Fleiss’s kappa, intra-class correlation)
Estimation of intra-rater, inter-rater, and test-retest reliability
Bland-Altman plots
Miscellaneous
Sample size calculations
Meta-analyses of results from the literature (plots, meta-regression)
Variable selection (lasso, elastic net)
Multivariate methods
High-dimensional analyses using copulas
