Statistical Thinking
Posts
Statistical Computing Approaches to Maximum Likelihood Estimation
Adjudication and Statistical Efficiency
The Burden of Demonstrating Statistical Validity of Clusters
Traditional Frequentist Inference Uses Unrealistic Priors
Borrowing Information Across Outcomes
Proportional Odds Model Power Calculations for Ordinal and Mixed Ordinal/Continuous Outcomes
The log-rank Test Assumes More Than the Cox Model
What Does a Statistical Method Assume?
Football Multiplicities
Incorporating Historical Control Data Into an RCT
Wedding Bayesian and Frequentist Designs Created a Mess
Ordinal Models for Paired Data
Randomized Clinical Trials Do Not Mimic Clinical Practice, Thank Goodness
Biostatistical Modeling Plan
How to Do Bad Biomarker Research
R Workflow
Decision curve analysis for quantifying the additional benefit of a new marker
Equivalence of Wilcoxon Statistic and Proportional Odds Model
Longitudinal Data: Think Serial Correlation First, Random Effects Second
Assessing the Proportional Odds Assumption and Its Impact
Commentary on Improving Precision and Power in Randomized Trials for COVID-19 Treatments Using Covariate Adjustment, for Binary, Ordinal, and Time-to-Event Outcomes
Incorrect Covariate Adjustment May Be More Correct than Adjusted Marginal Estimates
Avoiding One-Number Summaries of Treatment Effects for RCTs with Binary Outcomes
If You Like the Wilcoxon Test You Must Like the Proportional Odds Model
Violation of Proportional Odds is Not Fatal
RCT Analyses With Covariate Adjustment
Bayesian Methods to Address Clinical Development Challenges for COVID-19 Drugs and Biologics
Implications of Interactions in Treatment Comparisons
The Burden of Demonstrating HTE
Assessing Heterogeneity of Treatment Effect, Estimating Patient-Specific Efficacy, and Studying Variation in Odds ratios, Risk Ratios, and Risk Differences
Statistically Efficient Ways to Quantify Added Predictive Value of New Measurements
In Machine Learning Predictions for Health Care the Confusion Matrix is a Matrix of Confusion
Viewpoints on Heterogeneity of Treatment Effect and Precision Medicine
Musings on Multiple Endpoints in RCTs
Improving Research Through Safer Learning from Data
Is Medicine Mesmerized by Machine Learning?
Information Gain From Using Ordinal Instead of Binary Outcomes
How Can Machine Learning be Reliable When the Sample is Adequate for Only One Feature?
Bayesian vs. Frequentist Statements About Treatment Efficacy
Integrating Audio, Video, and Discussion Boards with Course Notes
EHRs and RCTs: Outcome Prediction vs. Optimal Treatment Selection
Statistical Errors in the Medical Literature
Damage Caused by Classification Accuracy and Other Discontinuous Improper Accuracy Scoring Rules
My Journey from Frequentist to Bayesian Statistics
A Litany of Problems With p-values
Clinicians’ Misunderstanding of Probabilities Makes Them Like Backwards Probabilities Such As Sensitivity, Specificity, and Type I Error
Split-Sample Model Validation
Fundamental Principles of Statistics
Classification vs. Prediction
Null Hypothesis Significance Testing Never Worked
p-values and Type I Errors are Not the Probabilities We Need
Talks
Bayesian Thinking
Modernizing Clinical Trial Design and Analysis to Improve Efficiency & Flexibility
Ordinal State Transition Models as a Unifying Risk Prediction Framework
Tips for Biostatisticians Collaborating with Non-Biostatistician Medical Researchers
Rare Degenerative Diseases & Statistics:
Methods for Analyzing Composite Patient Outcomes
Overview of Composite Outcome Scales & Statistical Approaches for Analyzing Them
My Big Jump: Founding a Department of Biostatistics
Controversies in Predictive Modeling, Machine Learning, and Validation
R Workflow for Reproducible Biomedical Research Using Quarto
Longitudinal Ordinal Models as a General Framework for Medical Outcomes
Musings on Statistical Models vs. Machine Learning in Health Research
Sequential Bayesian Designs for Rapid Learning in COVID-19 Clinical Trials
Bayes for Flexibility in Urgent Times
Fundamental Advantages of Bayes in Drug Development
R for Graphical Clinical Trial Reporting
Why Bayes for Clinical Trials?
R for Clinical Trial Reporting
Regression Modeling Strategies
Simple Bootstrap and Simulation Approaches to Quantifying Reliability of High-Dimensional Feature Selection
Exploratory Analysis of Clinical Safety Data to Detect Safety Signals
Courses
Regression Modeling Strategies Course
This course covers a comprehensive strategy for developing accurate predictive models, model specification that preserves information, quantifying predictive accuracy, avoiding overfitting, data reduction (unsupervised learning), making optimum use of incomplete data, validation, the art of data analysis, comprehensive case studies, and more. The course web site is here.
Regression Modeling Strategies Pre-Course
Even though the 4-day RMS course will not require you to use R interactively, those participants who wish to learn more about R or attain the regression knowledge prerequisite for the 4-day course may wish to take this optional one-day Pre-RMS workshop to enhance R and RStudio) skills, learn about multiple linear regression (a prerequisite for the 4-day course), and to get an introduction to the R rms package that will be used throughout the 4-day course. The course web site is here. The course introduces an R workflow for the entire analysis project cycle.