qscore_stat {crctStepdown}R Documentation

Calculates the randomisation test statistic

Description

Calculates a randomisation test statistic based on the sum of generalised, studentized residuals for a given model fit and null hypothesis of the value of the treatment effect parameter

Usage

qscore_stat(
  fit,
  data,
  null_par = 0,
  tr_var = "treat",
  cl_var = "cl",
  tr_assign = "treat"
)

Arguments

fit

A glm or lm model fit

data

The data frame used to fit model fit

null_par

Numeric the value of tr_var parameter under the null hypothesis

tr_var

String indicating the name of the column in data that is a binary indicator for whether the observation was under the treatment (1=treatment, 0=control)

cl_var

String specifying the name of the column identifying the clusters/cluster-time

tr_assign

String specifying the treatment assignment under a new permutation. If calculating the test statistics under the original treatment assignment then this should be the same as tr_var

Value

The value of the test statistic

Examples

out <- twoarm_sim()
data <- out[[1]]
  fit1 <- glm(y1 ~ treat ,
              data=data,
              family="poisson")
qscore_stat(fit=fit1,
            data=data)

[Package crctStepdown version 0.2.1 Index]