pairwise_t_test_mi: Pairwise t-tests() on multiply-imputed data (accepts survey...

In LukasWallrich/timesaveR: Functions to Accelerate (Academic) Data Analysis and Reporting

Pairwise t-tests() on multiply-imputed data (accepts survey weights)

Description

This runs pairwise t-tests (based on lm(), therefore assuming equal variances) on multiple imputations, with the ability to take into account survey weights.

Usage

pairwise_t_test_mi(
  mi_list,
  dv,
  groups,
  weights = NULL,
  p.adjust.method = p.adjust.methods
)

Arguments

mi_list	A list of dataframes, each consisting of a multiply imputed dataset
dv	The dependent variable for the t.test (must be in mi_list)
groups	The grouping variable (each distinct value will be treated as a level)
weights	The variable containing survey weights, must be in mi_list
p.adjust.method	The method to adjust p-values for multiple comparison (see stats::p.adjust())

Value

A tibble containing the results of the t-tests with one test per row

Examples

#Create list with imputed data
library(mice)
library(dplyr)
imp <- mice(nhanes2)
imp_list <- complete(imp, action="long") %>%
   group_split(.imp)

#Specify dependent variable and grouping factor
pairwise_t_test_mi(imp_list, bmi, age)

LukasWallrich/timesaveR documentation built on Nov. 29, 2024, 4:47 a.m.