Practical, careful guidance for planning, measurement, modeling, validation, and reproducible quantitative work.
Search by question, method, model, or research topic.
79 decision guides
Principal component analysis and factor analysis can produce similar-looking output while answering different research questions. This synthetic case study uses the same 12 correlated survey variables to show why PCA is a dimensionality-reduction method and common factor analysis is a latent-variable framework.
Read guideCollider bias shows why adding more variables to a causal regression model can increase rather than reduce bias. Learn how colliders, descendants, selection, participant restriction, and loss to follow-up can open noncausal paths—and how DAG-based reasoning helps identify variables that should and should not be adjusted for.
Read guideAn EFA should do more than produce a rotated loading matrix. This synthetic 18-item case study shows how construct definitions, correlations, factor retention, rotation, cross-loadings, and item content combine to produce a provisional factor structure—and why that structure still requires independent confirmation.
Read guideWhat does it mean when a regression result is “adjusted for” other variables? This Resource explains covariates, control variables, and confounders, how adjustment changes coefficient interpretation, and why an adjusted association is not automatically a causal effect.
Read guideA practical framework for choosing a statistical test in health and medical research based on the research question, outcome type, dependence structure, number of groups, target of inference, and method assumptions. It also explains when a simple test should give way to regression, survival analysis, or another modeling framework.
Read guidePropensity scores can improve comparability on measured pretreatment covariates and reveal poor overlap in observational studies. This Resource explains what matching, stratification, and weighting can accomplish—and why they cannot establish randomization or eliminate unmeasured confounding.
Read guideLearn how to distinguish cohort, case-control and cross-sectional studies by participant selection, time structure and the quantities each design can legitimately estimate, while keeping prospective versus retrospective timing conceptually separate from the underlying study design.
Read guideHaving several dependent variables does not automatically mean you should use MANOVA. This practical guide explains when a joint multivariate test answers the research question, when separate ANOVAs are more interpretable, how outcome correlations and Type I error affect the choice, and why repeated measurements are a different design problem.
Read guideMain effects summarize differences across a factor, while interactions ask whether those differences depend on another factor. Four synthetic 2 × 2 examples show why cell means, simple effects, and interaction plots matter alongside factorial ANOVA tests.
Read guidePoor CFA model fit does not mean you should keep following modification indices until CFI or RMSEA reaches a preferred value. This synthetic case study shows how to diagnose loadings, residuals, and modification evidence, distinguish theory-supported revisions from model fishing, and validate a revised confirmatory factor analysis.
Read guideRisk difference, risk ratio, and odds ratio describe different aspects of binary outcomes and should not be treated as interchangeable. This Resource explains how the estimand, study design, model, baseline risk, and communication goal should guide effect-measure selection and reporting.
Read guideA high correlation between two measurement methods does not establish that they agree. This Resource explains how paired differences, average bias, Bland-Altman limits of agreement, graphical assessment, repeatability, and diagnostic accuracy address distinct methodological questions.
Read guideYour design, measurement, data structure, assumptions, and existing output can change the appropriate statistical decision. If the choice is still unclear, discuss the study in context.