ggeffects (version 2.0.0)

vcov: Calculate variance-covariance matrix for adjusted predictions

Description

Returns the variance-covariance matrix for the predicted values from object.

Usage

# S3 method for ggeffects
vcov(object, vcov = NULL, vcov_args = NULL, verbose = TRUE, ...)

Value

The variance-covariance matrix for the predicted values from object.

Arguments

object

An object of class "ggeffects", as returned by predict_response().

vcov

Variance-covariance matrix used to compute uncertainty estimates (e.g., for confidence intervals based on robust standard errors). This argument accepts a covariance matrix, a function which returns a covariance matrix, or a string which identifies the function to be used to compute the covariance matrix.

• A covariance matrix

• A function which returns a covariance matrix (e.g., stats::vcov())

• A string which indicates the kind of uncertainty estimates to return.

  • Heteroskedasticity-consistent: "HC", "HC0", "HC1", "HC2", "HC3", "HC4", "HC4m", "HC5". See ?sandwich::vcovHC

  • Cluster-robust: "vcovCR", "CR0", "CR1", "CR1p", "CR1S", "CR2", "CR3". See ?clubSandwich::vcovCR.

  • Bootstrap: "BS", "xy", "fractional", "jackknife", "residual", "wild", "mammen", "norm", "webb". See ?sandwich::vcovBS

  • Other sandwich package functions: "HAC", "PC", "CL", or "PL".

If NULL, standard errors (and confidence intervals) for predictions are based on the standard errors as returned by the predict()-function. Note that probably not all model objects that work with predict_response() are also supported by the sandwich or clubSandwich packages.

See details in this vignette.

vcov_args

List of arguments to be passed to the function identified by the vcov argument. This function is typically supplied by the sandwich or clubSandwich packages. Please refer to their documentation (e.g., ?sandwich::vcovHAC) to see the list of available arguments. If no estimation type (argument type) is given, the default type for "HC" equals the default from the sandwich package; for type "CR" the default is set to "CR3". For other defaults, refer to the documentation in the sandwich or clubSandwich package.

verbose

Toggle messages or warnings.

...

Currently not used.

Details

The returned matrix has as many rows (and columns) as possible combinations of predicted values from the predict_response() call. For example, if there are two variables in the terms-argument of predict_response() with 3 and 4 levels each, there will be 3*4 combinations of predicted values, so the returned matrix has a 12x12 dimension. In short, nrow(object) is always equal to nrow(vcov(object)). See also 'Examples'.

Examples

data(efc)
model <- lm(barthtot ~ c12hour + neg_c_7 + c161sex + c172code, data = efc)
result <- predict_response(model, c("c12hour [meansd]", "c161sex"))

vcov(result)

# compare standard errors
sqrt(diag(vcov(result)))
as.data.frame(result)

# only two predicted values, no further terms
# vcov() returns a 2x2 matrix
result <- predict_response(model, "c161sex")
vcov(result)

# 2 levels for c161sex multiplied by 3 levels for c172code
# result in 6 combinations of predicted values
# thus vcov() returns a 6x6 matrix
result <- predict_response(model, c("c161sex", "c172code"))
vcov(result)