p_2r: p-value from comparing two or more correlations simulation

In Spower: Power Analyses using Monte Carlo Simulations

p-value from comparing two or more correlations simulation

Description

Function utilizes cocor to perform correlation comparison for independent, overlapping, and non-overlapping correlation designs. Type type of correlation design is inferred based on which correlations are specified.

Usage

p_2r(
  n,
  r.ab,
  r.ab2 = NULL,
  r.ac,
  r.bc,
  r.ad,
  r.bd,
  r.cd,
  n2_n1 = 1,
  two.tailed = TRUE,
  test = NULL,
  gen_fun = gen_2r,
  return_analysis = FALSE,
  ...
)

gen_2r(n, R, ...)

Arguments

n	sample size
r.ab	correlation between variable A and B (for independent groups, this is for sample 1)
r.ab2	(for independent group test only) correlation between variable A and B in sample 2
r.ac	(for overlap/non-overlap) correlation between A and C. This is the correlation used in the overlapping hypothesis test, comparing this input to r.ab
r.bc	(for overlap/non-overlap only) correlation between B and C.
r.ad	(for non-overlap only) correlation between A and D
r.bd	(for non-overlap only) correlation between B and D
r.cd	(for non-overlap only) correlation between C and D. This is the correlation used in the non-overlapping hypothesis test, comparing this input to r.ab
n2_n1	sample size ratio. Only used for independent group test
two.tailed	logical; use two-tailed test?
test	hypothesis testing method to use. Defaults to 'fisher1925' for the independent groups test and 'steiger1980' for overlap/non-overlap tests
gen_fun	function used to generate the required discrete data. Object returned must be a matrix with n rows. Default uses gen_2r. User defined version of this function must include the argument ...
return_analysis	logical; return the analysis object for further extraction and customization?
...	additional arguments to be passed to gen_fun. Not used unless a customized gen_fun is defined
R	a correlation matrix constructed from the inputs to p_2r

Details

For independent group tests, only r.ab and r.ab2 need to be specified, where the null hypothesis pertains to H_0: r_{ab}=r_{ab2}.

For overlapping correlation tests, r.ab, r.ac, and r.bc need to be specified, where the null hypothesis pertains to H_0: r_{ab}=r_{ac}.

For non-overlapping correlation tests, all correlations expect for r.ab2 must be specified, where the null hypothesis pertains to H_0: r_{ab}=r_{cd}.

Value

a single p-value

Author(s)

Phil Chalmers [e-mail]

Examples

# independent (same x-y pairing across groups)
p_2r(100, r.ab=.5, r.ab2=.6)

# return cocor object for further analysis
p_2r(100, r.ab=.5, r.ab2=.6, return_analysis = TRUE)



   # estimate empirical power
   p_2r(n=100, r.ab=.5, r.ab2=.6) |> Spower()

   # estimate n required to reach 80% power
   p_2r(n=NA, r.ab=.5, r.ab2=.6) |>
        Spower(power=.80, interval=c(100, 5000))



# overlap (same y, different xs)
# H0: r.ab = r.bc
p_2r(100, r.ab=.5, r.ac=.3, r.bc=.2)

# nonoverlap (different ys, different xs)
# H0: r.ab = r.cd
p_2r(100, r.ab=.5, r.ac=.3, r.bc=.2, r.ad=.2, r.bd=.4, r.cd=.2)

Spower documentation built on Nov. 28, 2025, 1:06 a.m.