Calibrate the KL normalized power prior to a design

The expected target standard error is the one the design implies, from npp_kl_expected_target_se(). It is not the standard error of any replicate: those vary around this one, and using them would make the prior a function of the data it is supposed to be a prior for.

Usage

npp_kl_calibrate_design(source, target_data, theta_0, settings)

Arguments

source

The source data, with treatment_effect_estimate and standard_error.

target_data

Target study data for the scenario.

theta_0

Boundary of the null hypothesis space.

settings

Output of npp_kl_settings().

Value

The output of calibrate_npp_kl().