Performs Wald or score tests
modelsearch(x, k = 1, dir = "forward", type = "all", ...)lvmfit-object
Number of parameters to test simultaneously. For equivalence
the number of additional associations to be added instead of rel.
Direction to do model search. "forward" := add associations/arrows to model/graph (score tests), "backward" := remove associations/arrows from model/graph (wald test)
If equal to 'correlation' only consider score tests for covariance parameters. If equal to 'regression' go through direct effects only (default 'all' is to do both)
Additional arguments to be passed to the low level functions
Matrix of test-statistics and p-values
m <- lvm();
regression(m) <- c(y1,y2,y3) ~ eta; latent(m) <- ~eta
regression(m) <- eta ~ x
m0 <- m; regression(m0) <- y2 ~ x
dd <- sim(m0,100)[,manifest(m0)]
e <- estimate(m,dd);
modelsearch(e,messages=0)
#> Score: S P(S>s) Index holm BH
#> 3.152 0.07584 y3~~x 0.5065 0.07584
#> 3.152 0.07584 y3~x 0.5065 0.07584
#> 3.152 0.07584 x~y3 0.5065 0.07584
#> 3.152 0.07584 y1~~y2 0.5065 0.07584
#> 3.152 0.07584 y1~y2 0.5065 0.07584
#> 3.152 0.07584 y2~y1 0.5065 0.07584
#> 4.127 0.04221 y1~~x 0.5065 0.06332
#> 4.127 0.04221 y1~x 0.5065 0.06332
#> 4.127 0.04221 x~y1 0.5065 0.06332
#> 4.127 0.04221 y2~~y3 0.5065 0.06332
#> 4.127 0.04221 y2~y3 0.5065 0.06332
#> 4.127 0.04221 y3~y2 0.5065 0.06332
#> 23.48 1.26e-06 y2~~x 2.269e-05 3.781e-06
#> 23.48 1.26e-06 y2~x 2.269e-05 3.781e-06
#> 23.48 1.26e-06 x~y2 2.269e-05 3.781e-06
#> 23.48 1.26e-06 y1~~y3 2.269e-05 3.781e-06
#> 23.48 1.26e-06 y1~y3 2.269e-05 3.781e-06
#> 23.48 1.26e-06 y3~y1 2.269e-05 3.781e-06
modelsearch(e,messages=0,type="cor")
#> Score: S P(S>s) Index holm BH
#> 3.152 0.07584 y3~~x 0.1688 0.07584
#> 3.152 0.07584 y1~~y2 0.1688 0.07584
#> 4.127 0.04221 y1~~x 0.1688 0.06332
#> 4.127 0.04221 y2~~y3 0.1688 0.06332
#> 23.48 1.26e-06 y2~~x 7.562e-06 3.781e-06
#> 23.48 1.26e-06 y1~~y3 7.562e-06 3.781e-06