itemStability: Item Stability Statistics from 'bootEGA'

Description Usage Arguments Value Author(s) References See Also Examples

View source: R/itemStability.R

Description

Based on the bootEGA results, this function computes and plots the number of times an item (variable) is estimated in the same factor/dimension as originally estimated by EGA (item.replication). The output also contains each item's replication frequency (i.e., proportion of bootstraps that an item appeared in each dimension; item.dim.rep) as well as the average network loading for each item in each dimension (item.loadings).

Usage

1
itemStability(bootega.obj, orig.wc, item.freq = 0.1, plot.item.rep = TRUE)

Arguments

bootega.obj

A bootEGA object

orig.wc

Numeric or character. A vector with community numbers or labels for each item. Typically uses community results (wc) from EGA

item.freq

A value for lowest frequency allowed in item.dim.rep output. Removes noise from table to allow for easier interpretation. Defaults to .10

plot.item.rep

Should the plot be produced for item.replication? If TRUE, then a plot for the item.replication output will be produced. Defaults to TRUE

Value

Returns a list containing:

item.replication

The proportion of times each item replicated within the defined dimension

mean.dim.rep

The average replication proportion of items replicating in each dimension. More simply, the average of the item.replication output for each dimension

item.dim.rep

The proportion of times each item replicated within each possible dimension. Dimensions greater than the maximum number used in the orig.wc argument are labeled based on the largest remaining components after the dimensions used to orig.wc

item.loadings

Matrix of the average standardized network loading (computed using net.loads) for each item in each dimension

wc

A matrix containing the community membership values for each bootstrapped sample. The values correspond to the values input for the orig.wc argument

plot.itemStability

A plot of the number of times each item replicates in its original community membership (orig.wc)

Author(s)

Hudson Golino <hfg9s at virginia.edu> and Alexander P. Christensen <alexpaulchristensen@gmail.com>

References

Christensen, A. P., & Golino, H. (2019). Estimating the stability of the number of factors via Bootstrap Exploratory Graph Analysis: A tutorial. PsyArXiv. doi: 10.31234/osf.io/9deay

Christensen, A. P., Golino, H., & Silvia, P. J. (2020). A psychometric network perspective on the validity and validation of personality trait questionnaires. European Journal of Personality. doi: 10.1002/per.2265

See Also

EGA to estimate the number of dimensions of an instrument using EGA and CFA to verify the fit of the structure suggested by EGA using confirmatory factor analysis.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
# Load data
wmt <- wmt2[,7:24]

# Estimate EGA network
## plot.type = "qqraph" used for CRAN checks
## plot.type = "GGally" is the default
ega.wmt <- EGA(data = wmt, model = "glasso", plot.type = "qgraph")

# Estimate dimension stability
boot.wmt <- bootEGA(data = wmt, iter = 100, typicalStructure = TRUE,
plot.typicalStructure = TRUE, model = "glasso", plot.type = "qgraph",
type = "parametric", ncores = 2)


# Estimate item stability statistics
res <- itemStability(boot.wmt, orig.wc = ega.wmt$wc)

# Changing plot features (ggplot2)
## Changing colors (ignore warnings)
### qgraph Defaults
res$plot.itemStability + 
    ggplot2::scale_color_manual(values = rainbow(max(res$uniq.num)))

### Pastel
res$plot.itemStability + 
    ggplot2::scale_color_brewer(palette = "Pastel1")
    
## Changing Legend (ignore warnings)
res$plot.itemStability + 
    ggplot2::scale_color_discrete(labels = "Intelligence")

EGAnet documentation built on Feb. 17, 2021, 1:06 a.m.