jtools: get_robust_se – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

get_robust_se

Calculate robust standard errors and produce coefficient tables

This function wraps around several sandwich and lmtest functions to calculate robust standard errors and returns them in a useful format.

get_robust_se(
  model,
  type = "HC3",
  cluster = NULL,
  data = model.frame(model),
  vcov = NULL
)

`model`	A regression model, preferably of class `lm` or `glm`
`type`	One of `"HC3"`, `"const"`, `"HC"`, `"HC0"`, `"HC1"`, `"HC2"`, `"HC4"`, `"HC4m"`, `"HC5"`. See `sandwich::vcovHC()` for some more details on these choices. Note that some of these do not work for clustered standard errors (see sandwich::vcovCL()]).
`cluster`	If you want clustered standard errors, either a string naming the column in `data` that represents the clusters or a vector of clusters that is the same length as the number of rows in `data`.
`data`	The data used to fit the model. Default is to just get the `model.frame` from `model`.
`vcov`	You may provide the variance-covariance matrix yourself and this function will just calculate standard errors, etc. based on that. Default is NULL.

A list with the following:

coefs: a coefficient table with the estimates, standard errors, t-statistics, and p-values from lmtest.
ses: The standard errors from coefs.
ts: The t-statistics from coefs.
ps: The p-values from coefs.
type: The argument to robust.
use_cluster: TRUE or FALSE indicator of whether clusters were used.
cluster: The clusters or name of cluster variable used, if any.
vcov: The robust variance-covariance matrix.

Analysis and Presentation of Social Scientific Data

v2.1.3

MIT + file LICENSE

Authors

Jacob A. Long [aut, cre] (<https://orcid.org/0000-0002-1582-6214>)

Initial release