Clinical Research Methods and Statistics

Learn how a medical manuscript is structured to present research clearly, from title and abstract through introduction, methods, results, and discussion, with emphasis on transparency and reproducibility.

Investigate the replication crisis and how selective publication, lack of transparency, and open data sharing affect reproducibility and trust in clinical research.

Classify data as categorical, continuous, ordinal, or qualitative to guide analysis and presentation in medical research, and apply statistical tools such as chi-squared tests, descriptive statistics, t tests, and Anova.

Explore central tendency and dispersion through mean, median, and mode, and examine range, standard deviation, and IQR to interpret variability in medical data.

Explore how sampling and statistics enable generalizable research findings by using appropriate sample sizes, methods, and measures of error and validity.

Explore the layers of medical data from simple counts to ratios and ratios of ratios, using relative risk, hazard ratios, and odds ratios to inform clinical decisions and public health.

Compare relative risk and absolute risk to interpret treatment outcomes and avoid misinterpretation. Learn how the number needed to treat quantifies real-world benefit and informs ethical decisions in medicine.

Explore how the p value measures evidence against the null hypothesis, distinguishes statistical from clinical significance, and how sample size, statistical power, and effect size influence interpretation.

Select appropriate statistical tests based on data type and distribution, conduct hypothesis testing with null and alternative hypotheses, and interpret p values to draw valid conclusions.

Understand negative and underpowered studies and how proper sample size and study design yield credible conclusions. Explore how statistical power, confidence intervals, mcid, and mrd shape interpretation and decision making.

Explore how randomized controlled trials establish cause and effect by using randomization, controls, and blinding to test medical interventions and establish gold-standard evidence.

Explore how cohort studies, an observational research design, track exposed and unexposed groups over time to assess exposures and health outcomes, identifying risk factors and long-term disease links.

Compare cases and controls in a retrospective design to uncover potential exposures and risk factors linked to a health outcome, using odds ratios and bias considerations.

Explore how diagnostic tests identify diseases using imaging, labs, and biopsies, and interpret sensitivity, specificity, and positive and negative predictive values in the context of prevalence.

Explore how causality links risk factors to disease incidence and prevalence, using cross-sectional counterfactuals, Bradford Hill criteria, and confounding controls.

Identify how confounding variables distort the relationship between exposure and outcome, threaten internal validity, and how stratification, regression, and propensity scores mitigate bias.

Explore how effect modification and mediation reveal causal pathways in treatment effects across subgroups. Identify mediators and confounders, and use regression with interactions to quantify direct and indirect effects.

Examine how selection bias arises from nonrepresentative samples, the difference between bias and confounding, and how trial design and intention-to-treat analysis mitigate bias in clinical research.

Explain how information bias distorts study findings through systematic mismeasurement across groups, including detection, recall, self-report, and co intervention biases.

Examine time period bias, immortal time bias, lead time bias, and length time bias in clinical research, and apply time varying analyses and baseline classifications.

Apply adjustment in research to account for confounders and isolate the exposure effect on outcomes. Use stratification, regression, and propensity score matching to reduce bias and interpret adjusted results.

Explore how matching reduces bias in observational studies by balancing covariates between treated and untreated groups, using exact, caliper, and propensity score matching to estimate true effects.

Propensity scoring estimates treatment effects by balancing covariates to reduce confounding in observational studies, using matching, stratification, or weighting to resemble randomized trials.