Bf_samplesize: Calculate Sample Size for Bayes Factor

In BencePalfi/SampleSizeBF: PKG_TITLE

Calculate Sample Size for Bayes Factor

Description

This function calculates the minimum sample size required to obtain evidence for H1 or H0 using Bayes factors. It supports different likelihood and prior distributions and assumes equal group sizes for between-subject designs.

Usage

Bf_samplesize(
  distribution_likelihood = c("normal", "t"),
  distribution_prior = c("normal", "cauchy", "uniform"),
  sd_of_theory,
  tail = c(1, 2),
  n_start,
  n_end,
  step,
  sd1,
  sd2 = NULL,
  threshold = c(3, 6, 10),
  tpr = c(0.5, 0.8, 0.9, 0.95)
)

Arguments

distribution_likelihood	Distribution for the likelihood function ("normal" or "t").
distribution_prior	Distribution for the prior ("normal", "cauchy", or "uniform").
sd_of_theory	The standard deviation of the theoretical distribution.
tail	The number of tails (1 or 2).
n_start	Starting sample size.
n_end	Maximum sample size.
step	Step size for sample size increments.
sd1	Standard deviation of the first sample.
sd2	Standard deviation of the second sample (for between-subject designs).
threshold	Bayes factor threshold for stopping.
tpr	True positive rate.

Value

A list with required sample sizes for H1 and H0.

BencePalfi/SampleSizeBF documentation built on April 3, 2025, 2:01 p.m.