Topic: Treatment-Effect Estimation After Selection

Worked Example: Combining Two Stages After Selection

Worked Example: Correcting a Selected Arm in a Two-Arm Trial

Worked Example: Three Arms With Interim Dropping

Why Naive Bootstrap Corrections Fail at Selection Boundaries

Why Naive Estimates Exaggerate Selected Treatment Effects

Tuning Shrinkage When Arms Were Selected Adaptively

Unequal Allocation and Post-Selection Estimation

Using a Validation Stage to Estimate Selected Effects

Using Later-Stage Outcomes After Interim Selection

The Selected Arm and the Truly Best Arm