Learn.

Learn statistical thinking and data analysis through the lens of music therapy research. No calculus or advanced math background required. Progress from descriptive statistics to regression and longitudinal analysis using real and simulated MT datasets.

The sequel to Statistics for Music Therapy Research. Covers advanced methods that MT researchers encounter in the literature and need for doctoral work: scale validation and psychometrics, structural equation modeling, Bayesian reasoning, advanced mixed models, meta-analysis, and writing a statistical analysis plan for grant proposals. Prerequisite: completion of the introductory MT statistics course or equivalent knowledge.

A plain-English guide to placebos, control groups, and blinding in clinical research. Designed for clinicians, nursing students, and anyone who reads medical literature but wants to understand why blinded trials produce more trustworthy results — with just enough statistics to see why bias is so dangerous.

Statistical significance tells you something is probably real. Effect size tells you whether it matters. This course covers Cohen's d, r, η², odds ratios, risk ratios, and Number Needed to Treat — transforming how you read and design research. Essential for anyone who publishes, reviews, or acts on scientific findings.

Walk through the complete process of designing a clinical trial: framing your research question, choosing a design, calculating sample size, building a randomization scheme, writing an analysis plan, and understanding regulatory expectations. Practical and accessible for clinicians and health researchers.

Master the statistical methods used whenever the outcome is 'how long until something happens' — death, relapse, discharge, device failure. Covers Kaplan-Meier curves, log-rank tests, and Cox proportional hazards regression using real clinical datasets, with no calculus required.

Most clinicians skip the statistics section of papers they read. This course changes that. Learn to decode every element of a published biostatistics results section — tables, forest plots, confidence intervals, regression coefficients — so you can evaluate evidence on your own terms.

A comprehensive statistics course built around food science data. Covers experimental design for sensory panels, ANOVA for formulation studies, correlation and regression for physicochemical properties, and a thorough introduction to principal component analysis (PCA) — the gateway to chemometrics. Real food datasets throughout.

The advanced continuation of Part 1. Master Partial Least Squares regression (PLS), PLS Discriminant Analysis (PLS-DA), model validation strategies, and variable importance methods. Applied to NIR spectra, NMR fingerprinting, and volatile compound profiles. Covers authentication of food products, fraud detection, and building robust predictive models.

Learn to conduct a systematic review and meta-analysis from start to finish. Covers literature searching, data extraction, forest plots, heterogeneity assessment, publication bias, and subgroup analyses. Uses the R meta package throughout. Perfect for researchers wanting to synthesize a body of evidence.

Bayesian statistics offers a more intuitive way to reason about uncertainty: you start with what you know, collect data, and update. This course builds deep Bayesian intuition using JASP and visual simulations before introducing any formulas. Perfect for researchers frustrated by the limitations of p-values.

Most surveys are poorly designed — and most Likert data is analyzed incorrectly. This course covers question writing, response scale design, reliability (Cronbach's α), validity, and the great Likert debate: treat it as ordinal or interval? Practical and opinionated, grounded in psychometric theory.

Most research is not reproducible. This course teaches you to structure your R projects, write literate analysis documents in Quarto, manage dependencies, version-control your code with Git, and archive your data so anyone can reproduce your findings from scratch. A skill set that reviewers, collaborators, and hiring committees increasingly demand.

Missing data is everywhere in research, and deleting incomplete rows is almost never the right answer. This course covers the missing data mechanisms (MCAR, MAR, MNAR), multiple imputation, sensitivity analysis, and how to report your approach transparently — using R's mice package throughout.

Patients nested in hospitals. Students nested in classrooms. Repeated measures nested in people. Standard regression assumes independence — multilevel (mixed-effects) models handle the reality. This course builds intuition from the ground up, then implements models in R's lme4 package with real nested data.

Does anxiety mediate the effect of music therapy on pain? Does treatment effectiveness depend on patient personality? This course covers mediation analysis, moderation (interaction effects), and full SEM with latent variables — using the lavaan R package with real behavioral and clinical datasets.

Zero to functional R in four weeks. Learn to navigate RStudio, import data, wrangle it with the tidyverse, visualize it with ggplot2, and run your first statistical tests — all with no prior programming experience. Designed for researchers who have avoided R long enough.

Most research figures are cluttered, misleading, or just ugly. This course teaches the principles of effective scientific visualization and the R tools to execute them. Covers chart selection, color theory for accessibility, publication standards, and interactive graphics — building toward figures that journals actually want.

Underpowered studies are everywhere — and they waste resources, time, and patient goodwill. This course builds deep intuition for statistical power from the ground up, then walks through sample size calculations for every common research design using G*Power and R. A must for anyone writing a grant or IRB protocol.

Most health research is observational — you can't randomly assign people to smoke or be obese. This course covers the major observational designs (cohort, case-control, cross-sectional, ecological), their statistical analyses, and the critical methods for reducing confounding: matching, restriction, stratification, and propensity scores.

Anxiety, pain, quality of life, personality — these are latent constructs we measure indirectly through questionnaire items. This course covers exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and the psychometric principles needed to develop or validate a measurement scale from scratch.

Health data is often collected over time — daily infections, weekly therapy adherence, monthly readmissions. This course covers time series decomposition, autocorrelation, ARIMA models, and interrupted time series analysis (ITS) for evaluating policy or clinical interventions without a randomized design.

When your outcome is binary — survived or died, passed or failed, infected or not — linear regression breaks down. This course teaches logistic regression from intuition to implementation: odds, log-odds, the logit link, model fitting, interpretation, and diagnostics. Hands-on with R throughout.

The bootstrap is one of the most powerful ideas in modern statistics: resample your data with replacement thousands of times and watch the sampling distribution emerge. This course covers the nonparametric bootstrap, bootstrap confidence intervals, permutation tests, and cluster bootstrap — all in R.

Simulation is how statisticians test whether methods actually work. This course teaches you to design, run, and analyze Monte Carlo simulation studies in R — from generating data under known conditions to measuring bias, variance, coverage, and power. Essential for anyone developing or validating statistical methods.

Bayesian networks represent the joint probability distribution of many variables as a directed acyclic graph (DAG). This course covers conditional independence, d-separation, structure learning, parameter estimation, and simulation from fitted networks — using the bnlearn R package. Ideal preparation for causal inference and missing data research.

Chi-square tests are just the beginning. When your categorical data has ordered categories, multiple dimensions, or longitudinal structure, you need more powerful tools. This course covers ordinal logistic regression, log-linear models, Cramer's V, and methods for analyzing multi-way contingency tables in R.

Correlation is not causation, but causation IS what we want. This course introduces the potential outcomes framework, the fundamental problem of causal inference, and the methods statisticians use to estimate causal effects from observational data: matching, inverse probability weighting, instrumental variables, and difference-in-differences.

Linear regression assumes a continuous, normally distributed outcome. But what about counts, rates, proportions, or strictly positive values? Generalized Linear Models (GLMs) extend regression to handle all of these through a unified framework of link functions and exponential family distributions. This course covers Poisson, negative binomial, gamma, and beta regression in R.

Every diagnostic test makes errors. This course teaches you to quantify those errors: sensitivity, specificity, positive and negative predictive values, likelihood ratios, and ROC curves. Essential for clinicians interpreting test results, researchers validating screening instruments, and anyone building classification models.

When your outcome variable is a proportion (0 to 1), percentage, or rate, linear regression can predict impossible values. Beta regression respects the bounded nature of your data, models both the mean and the precision, and handles the asymmetric distributions that proportions naturally produce. This short course gets you up and running with the betareg R package.

Shiny lets you build interactive web applications entirely in R. This course takes you from zero to a deployed dashboard: reactive inputs, dynamic plots, data tables, and publishing to shinyapps.io. Perfect for researchers who want collaborators to explore their data without touching R.

A guided research project course. Most simulation approaches generate complete data then overlay an independent missingness mechanism — but real missingness is often MNAR (depends on unobserved values). This course walks you through a Bayesian network approach to learning and reproducing the joint dependency structure between categorical outcomes and their missingness indicators from real data. You will build the full pipeline: seed data generation, network structure learning, parameter estimation, simulation, and validation.

A guided research project course. Among patients who underwent open aneurysm repair and experienced a Major Adverse Event, how many EVAR patients have identical covariate profiles — and what proportion also had an MAE? This course walks you through a complete exact-matching simulation study: synthetic data generation, match key construction, 1:1 vs. all-matches designs, Wilson score and cluster-bootstrap confidence intervals, Monte Carlo evaluation, and clinical interpretation.

A guided research project course. When group survival depends on the balance of a trait among its members, can statistical methods detect this relationship? This course walks you through a multi-method comparative study: simulating group survival data with known balance effects, then analyzing it with chi-squared goodness-of-fit, beta regression, and binomial GLMs. You will learn why different methods suit different data structures.

Differential equations are the language of dynamics — they describe how things change over time and space. This course provides a thorough treatment of ordinary and partial differential equations: separable and exact equations, second-order linear ODEs, systems of equations, Laplace transforms, series solutions, numerical methods (Euler, Runge-Kutta), and an introduction to heat, wave, and Laplace equations. Theory is paired with R implementations throughout.

Real analysis provides the rigorous mathematical foundations underlying all of calculus and modern analysis. This course covers the completeness of the real numbers, sequences and series, topology of R, continuity and uniform continuity, differentiation, Riemann integration, sequences and series of functions, power series, and an introduction to metric spaces and measure theory. Proofs and intuition are developed hand-in-hand, with R demonstrations to build numerical intuition.