| sbs.log {SBSDiff} | R Documentation |
Satorra-Bentler Scaled Chi-Squared Difference Test (Based on Loglikelihood Values)
Description
Takes loglikelihood values from nested models estimated using maximum likelihood with robust standard errors, number of free parameters, scaling correlation factors and returns: (1) Satorra-Bentler scaled change in model chi-squared (2) change in model degrees of freedom and (3) the probability of rejecting the null.
Usage
sbs.log(L0, L1, p0, p1, c0, c1)
Arguments
L0 |
loglikelihood value for the more restrictive model (should be a negatige value) |
L1 |
loglikelihood value for the less restrictive model (should be a negatige value) |
p0 |
number of free parameters for the more restrictive model (with fewer freely estimated parameters) |
p1 |
number of free parametersfor the less restrictive model (with more freely estimated parameters) |
c0 |
scaling correction factor for the more restrictive model |
c1 |
scaling correction factor for the less restrictive model |
Value
Change in model chi-squared, change in model degrees of freedom and the probability of rejecting the null
Examples
L0 <- -50
L1 <- -45
p0 <- 9
p1 <- 10
c0 <- 1
c1 <- 1
sbs.log(L0,L1,p0,p1,c0,c1)