Skip to contents

Sample Size or Power for Odds Ratio Test

Source: R/or_size.R
or_size.Rd

Calculates sample size or power for odds ratio test.

Usage

or_size(
  pA,
  pB,
  delta = NULL,
  kappa = 1,
  alpha,
  beta = NULL,
  nB = NULL,
  test_type = "equality"
)

Arguments

pA

Numeric. True proportion of group A.

pB

Numeric. True proportion of group B.

delta

Numeric (optional). Margin for "non-inferiority" or "equivalence" test. Required for "non-inferiority" or "equivalence" test.

kappa

Numeric. Ratio of sample sizes (nA/nB). Default is 1.

alpha

Numeric. Type I error rate.

beta

Numeric (optional). Type II error rate. Required for sample size calculation.

nB

Integer (optional). Sample size for group B. Required for power calculation.

test_type

Character. "equality", "non-inferiority", or "equivalence". Default is "2-side".

Value

Numeric. Returns sample size (if beta is given), or power (if nB is given).

Note

Only one of beta (for sample size calculation) or nB (for power calculation) should be specified.

Required arguments by test_type:

  • "equality":

    • For sample size: pA, pB, alpha, beta

    • For power: pA, pB, alpha, nB

  • "non-inferiority"/"equivalence":

    • For sample size: pA, pB, delta, alpha, beta

    • For power: pA, pB, delta, alpha, nB

Examples

# Sample size for `"equality"` test
or_size(pA = 0.4, pB = 0.25, kappa = 1,
        alpha = 0.05, beta = 0.2, test_type = "equality")
#> [1] 156

# Power of `"equality"` test
or_size(pA = 0.4, pB = 0.25, kappa = 1,
        alpha = 0.05, nB = 156, test_type = "equality")
#> [1] 0.8020239

# Sample size for `"non-inferiority"` test
or_size(pA = 0.4, pB = 0.25, delta = 0.2, kappa = 1,
        alpha = 0.05, beta = 0.2, test_type = "non-inferiority")
#> [1] 242

# Power of `"non-inferiority"` test
or_size(pA = 0.4, pB = 0.25, delta = 0.2, kappa = 1,
        alpha = 0.05, nB = 242, test_type = "non-inferiority")
#> [1] 0.8007201

# Sample size for `"equivalence"` test
or_size(pA = 0.25, pB = 0.25, delta = 0.5, kappa = 1,
        alpha = 0.05, beta = 0.2, test_type = "equivalence")
#> [1] 366

# Power of `"equivalence"` test
or_size(pA = 0.25, pB = 0.25, delta = 0.5, kappa = 1,
        alpha = 0.05, nB = 366, test_type = "equivalence")
#> [1] 0.8008593