Title

commarobust(fit, clust_var = NULL, type = "HC2", alpha = 0.05)

Arguments

fit A model object, typically created by lm, glm, or many other model fitting functions. An optional numeric vector of length N that describes the clusters that units are in. May not include NAs. Is NULL by default. If specified, cluster-robust standard errors will be estimated. A string indicating the type of heterskedasticty-robust standard errors to be estimated. Is ignored if clust_var is specificed. Is "HC2" by default, because these are equivalent to Neyman standard errors (See Aronow and Samii). Used for construction of (1 - alpha) confidence intervals. Defaults to 0.05.

Value

A matrix of coefficients, standard errors, t-statistics, and p-values

Examples


library(randomizr)#> Warning: package ‘randomizr’ was built under R version 3.3.2
Y <- rnorm(100)
Z <- complete_ra(100)
fit <- lm(Y ~ Z)
commarobust(fit)#>               Estimate Std. Error   t value  Pr(>|t|)
#> (Intercept)  0.1996898  0.1444663  1.382259 0.1700345
#> Z           -0.2375446  0.2079651 -1.142233 0.2561403
# Clustered
Y <- rnorm(100)
clust_var <- rep(letters[1:10], 10)
Z <- cluster_ra(clust_var = clust_var)

fit <- lm(Y ~ Z)
commarobust(fit, clust_var = clust_var)#>                Estimate Std. Error     t value  Pr(>|t|)
#> (Intercept) -0.04796144  0.1234366 -0.38855121 0.6984510
#> Z           -0.00546281  0.1671615 -0.03267983 0.9739964