Variability tests

vartest performs variability tests by either the Breusch-Pagan or Brown-Forsythe methods.

Usage

vartest(y, x, covar = NULL, covar.var = FALSE, type = 1, x.sq = FALSE)

Arguments

y

vector of outcome values

x

vector of exposure values

covar

data.frame of covariates

covar.var

adjust the second stage (variance component) of the approach by the covariates

type

type of test, where 1 = Breusch-Pagan variance test, and 2 = Brown-Forsythe variance test (default: 1)

x.sq

include x-squared in the variance part of the model

Value

vartest returns a list of results:

coef

model coefficients from variance part of the model

test

data.frame of test results

• F: the test statistic

• DF: the degrees of freedom

• P: the p-value

Author

James Staley jrstaley95@gmail.com

Examples

x <- rbinom(1000, 1, 0.5)
y <- 0.5 + 0.025 * x + rnorm(1000, 0, sqrt(0.005 * x)) + rnorm(1000, 0, 0.1)
vartest(y, x, type = 2)
#> $coef
#>               Estimate  Std. Error   t value      Pr(>|t|)
#> (Intercept) 0.07900124 0.003082579 25.628290 1.030821e-111
#> x           0.01966774 0.004359426  4.511543  7.200682e-06
#> 
#> $test
#>          F DF            P
#> 1 20.35402  1 7.200682e-06
#>