omg: Run an OMG Analysis,

In clbustos/rcebn: Multiple methods for psychologist

Run an OMG Analysis,

Description

As Stuive, Kiers & Timmerman(2009) said,"The OMG method considers for each item the observed correlations between the item and each subtest, where each subtest is an unweighted sumscore of items. These correlations are corrected for two sources of spuriously high correlations, namely, self-correlation and test length[...] An assignment of an item to a subtest is considered to be correct when the corrected subtest-item correlation is highest for the subtest the item is supposed to belong to and incorrect otherwise" (p.950)

Usage

omg(x, keys, partialize = NULL, ...)

Arguments

x	matrix or data.frame
keys	keys of assigment of
...	arguments to cor.corrected

Value

a data.frame

item

Item name

assigned.name

Name of scale (came from keys)

cor.assigned

Corrected correlation of item to assigned scale

max.name

Name of scale for maximum correlation

cor.max

Maximum correlation to any scale

diff

Difference between assigned and maximum correlation.

See Also

cor.corrected for corrected correlation for self-correlation and scales lengths.

Examples

library(qgraph)
data(big5)
data(big5groups)
big.keys<-sapply(big5groups,function(x) {xx<-numeric(ncol(big5));xx[x]<-1;xx})
rownames(big.keys)<-colnames(big5)
omg(big5,big.keys)

clbustos/rcebn documentation built on Feb. 2, 2024, 5:55 a.m.