DepressionDemo: Artificial depression treatment dataset

Description

Simulated dataset of a randomized clinical trial (N = 150) to illustrate fitting of (G)LMM trees.

Usage

data("DepressionDemo")

Arguments

Format

A data frame containing 150 observations on 6 variables:

depression: numeric. Continuous treatment outcome variable (range: 3-16, M = 9.12, SD = 2.66).
treatment: factor. Binary treatment variable.
cluster: factor. Indicator for cluster with 10 levels.
age: numeric. Continuous partitioning variable (range: 18-69, M = 45, SD = 9.56).
anxiety: numeric. Continuous partitioning variable (range: 3-18, M = 10.26, SD = 3.05).
duration: numeric. Continuous partitioning variable (range: 1-17, M = 6.97, SD = 2.90).
depression_bin: factor. Binarized treatment outcome variable (0 = recovered, 1 = not recovered).

Details

The data were generated such that the duration and anxiety covariates characterized three subgroups with differences in treatment effects. The cluster variable was used to introduce a random intercept that should be accounted for. The treatment outcome is an index of depressive symptomatology.

Examples

Run this code

# \donttest{
data("DepressionDemo", package = "glmertree")
summary(DepressionDemo)
lt <- lmertree(depression ~ treatment | cluster | anxiety + duration + age, 
        data = DepressionDemo)
plot(lt)
gt <- glmertree(depression_bin ~ treatment | cluster | anxiety + duration + age, 
        data = DepressionDemo)
plot(gt)
# }