kutils (version 1.0)

semTable: Creates Structural Equation Modeling Tables

Description

Creates LaTeX markup for structural equation modeling output tables in the style of the American Psychological Association(APA). Input objects should be created by the "lavaan" package.

Usage

semTable(object, file = NULL, params = "all", fit = c("chi-square", "cfi",
  "tli", "rmsea"), names_fit = fit, standardized = FALSE,
  names_upper = TRUE, single_spaced = TRUE, type = "latex",
  group = NULL, longtable = FALSE)

Arguments

object

A lavaan object returned by cfa() or sem().

file

Character string for file name. Default is NULL, meaning no output file.

params

Parameters to be included. Valid values are "loadings", "slopes", "intercepts", "means", "residuals", "covariances", "latentvariances", "latentmeans" and "thresholds". Defaults to "all" which includes all available parameters (but excludes "means"). See Details.

fit

A vector of fit measures that to be included in the note. Listing "chi-square" will do special formatting to the chi-square value in the note. Any other measures listed must correspond to measures found in fitMeasures(object).

names_fit

Names for the fit measures requested by the fit parameter. Must have same number of elements as fit. For example, fit = c("cfi.scaled", "tli.scaled"), names_fit = c("CFI", "TLI").

standardized

Should standarized results be presented along with unstandardized? Default is FALSE. See Details.

names_upper

Should the names of the model fit parameters be forced to be uppercase. The default is TRUE. This will also affect whatever is specified in names_fit.

single_spaced

Default = TRUE. If a double-spaced table is needed, set single_spaced = FALSE.

type

Type of output table ("latex" or "html"). Defaults to "latex".

group

Group for which parameters should be reported. Provide the value in the data that indicates the desired group. Only necessary for multiple group models. Defaults to NULL.

longtable

Should a latex longtable be generated? Defaults to FALSE, which makes the table tabular. Ignored if type = "html".

Value

SEM table of desired type.

Details

The argument params determines the inclusion of estimate sections.

• "loadings" are the factor loadings in the model.

• "slopes" are the regression slopes in the model.

• "intercepts" are the observed variable intercepts.

• "means" are the observed variable means.

• "residuals" are the observed variable residual variances.

• "covariances" are the observed variable covariances.

• "latentvariances" are the latent variable variances and covariances.

• "latentmeans" are the latent variable means (or intercepts).

• "thresholds" arise in latent response variates (non-numeric indicator data).

The stantardized parameter regulates the number of columns to be included. standardized=FALSE implies there will be four columns:

1. the estimate

2. the standard error

3. the z-value, and

4. the p-value.

When standardized = TRUE, the columns are:

1. the parameter estimates,

2. the standard errors,

3. standardized parameter estimates, and

4. standardized standard errors.

The standardized parameters are obtained by updating the output with the options std.lv = TRUE and std.ov = TRUE. If these options were used when originally creating output, setting standardized = TRUE will yield two identical sets of two columns.

Examples

# NOT RUN {
## These run longer than 5 seconds
## CFA model
require(lavaan)
HS.model <- ' visual  =~ x1 + x2 + x3
textual =~ x4 + x5 + x6
speed   =~ x7 + x8 + x9'
output1 <- cfa(HS.model, data = HolzingerSwineford1939, std.lv = TRUE)
semTable(output1, fit = "rmsea",
standardized = TRUE, type = "latex")
## Basic SEM model
regmodel <- "x1 ~ x2 + x3
x1 ~1"
output1a <- sem(regmodel, data = HolzingerSwineford1939, std.lv = TRUE)
semTable(output1a, fit = "rmsea", type = "html")
#### Example with file output
##semTable(output1, file = "exampleTable.html", fit = "rmsea",
##standardized = TRUE, params = c("loadings", "latentvariances"),
##type = "html")

model <- "factor =~ .7*y1 + .7*y2 + .7*y3 + .7*y4
y1 | -1*t1 + 1*t2
y2 | -.5*t1 + 1*t2
y3 | -.2*t1 + 1*t2
y4 | -1*t1 + 1*t2
"
dat <- simulateData(model, sample.nobs = 300)
testmodel <- "ExampleFactor =~ y1 + y2 + y3 + y4"
output2 <- cfa(testmodel, data = dat, ordered = colnames(dat),
    std.lv = FALSE)
semTable(output2,
    params = c("loadings", "thresholds", "residuals"),
    fit = c("tli", "chi-square"),
    names_fit = c("TLI", "chi-square"), type = "html")

## Example with file output
## semTable(output, file = "catTable.tex",
##    params = c("loadings", "thresholds", "residuals"),
##    fit = c("tli", "chi-square"),
##    names_fit = c("TLI", "chi-square"), type = "latex")
# }