semPower (version 1.2.0)

semPower.aPriori: semPower.aPriori

Description

Determine required sample size given alpha, beta/power, df, and effect

Usage

semPower.aPriori(
  effect = NULL,
  effect.measure = NULL,
  alpha,
  beta = NULL,
  power = NULL,
  N = NULL,
  df,
  p = NULL,
  SigmaHat = NULL,
  Sigma = NULL
)

Value

list

Arguments

effect

effect size specifying the discrepancy between H0 and H1 (a list for multiple group models)

effect.measure

type of effect, one of "F0", "RMSEA", "Mc", "GFI", AGFI"

alpha

alpha error

beta

beta error; set either beta or power

power

power (1-beta); set either beta or power

N

a list of sample weights for multiple group power analyses, e.g. list(1,2) to make the second group twice as large as the first one

df

the model degrees of freedom

p

the number of observed variables, required for effect.measure = "GFI" and "AGFI"

SigmaHat

model implied covariance matrix (a list for multiple group models). Use in conjuntion with Sigma to define effect and effect.measure.

Sigma

population covariance matrix (a list for multiple group models). Use in conjuntion with SigmaHat to define effect and effect.measure.

Examples

if (FALSE) {
power <- semPower.aPriori(effect = .05, effect.measure = "RMSEA", alpha = .05, beta = .05, df = 200)
summary(power)
power <- semPower.aPriori(effect = .15, effect.measure = "F0", alpha = .05, power = .80, df = 100)
summary(power)
power <- semPower.aPriori(effect = list(.05, .10), effect.measure = "F0", alpha = .05, 
                          power = .80, N = list(1, 1), df = 100)
summary(power)
power <- semPower.aPriori(alpha = .01, beta = .05, df = 5, 
                          SigmaHat = diag(4), Sigma = cov(matrix(rnorm(4*1000),  ncol=4)))
summary(power)
}