Topic: Treatment Selection and Arm Dropping

Worked Selection Rule For Binary Outcomes

Worked Holm Analysis After Selecting Two Arms

Worked MaxT Analysis With Correlated Treatment Statistics

Why Winner Selection Inflates Unadjusted Type I Error

Winner’s Curse In Selected Treatment Effect Estimates

Worked Dunnett Analysis With Three Experimental Arms

Verifying Strong Control Across Null Configurations

Treatment Selection With A Shared Control

Threshold-Based Arm Retention And Dropping

Stress-Testing Correlated Outcomes And Unequal Effects