Skip to contents

Mirrors the arithmetic the simulation itself uses (R/simulation_scenarios.R): the total target sample size is the source study's arm sizes summed and divided by the factor, and the per-arm size is half of that, floored. Note this uses control + treatment rather than the total: field, which duplicates them and has drifted out of step before (aprepitant's once read 673).

Usage

paper_sample_size_per_arm(case_study, factor, case_studies_config_dir)

Arguments

case_study

Case study name.

factor

Sample size factor.

case_studies_config_dir

Directory holding the case study YAMLs, trailing slash included.

Value

The target sample size per arm, as an integer.