calc_power: Calculate power for cross-site variation

In lmiratrix/blkvar: ATE and Treatment Variation Estimation for Blocked and Multisite RCTs

Calculate power for cross-site variation

Description

Calculate power for detecting cross site variation with a ratio test. This uses work and results in Appendix A of Bloom and Spybrook.

Usage

calc_power(J, n.bar, tau, p.tx = 0.5, ICC = 0.5, K = 0, R2.X = 0, alpha = 0.05)

Arguments

J	Number of sites
n.bar	average people per site
tau	Cross site VARIANCE of site-level ATEs.
p.tx	proportion treated, Default: 0.5
ICC	Control-side ICC, Default: 0.5
K	Number of individual covariates used to explain control-side variation, Default: 0
R2.X	Explanatory power of these individual level covariates, Default: 0
alpha	Significance level, Default: 0.05

Value

Power

Examples

# Example in the text
calc_power( 80, 60, 0.02, p.tx=0.6, ICC=0.2, K = 0, R2.X=0.25 )

lmiratrix/blkvar documentation built on Nov. 18, 2024, 1:27 p.m.