create_regeq: Create mixed-effects regression formula from multi-level...

In jadahlke/hlmer: HLM7-Style Output from LME4 Analyses

Description

Create mixed-effects regression formula from multi-level formulas from up to 3 levels.

Usage

create_regeq(eq_lvl1, eq_lvl2, eq_lvl3 = NULL, cluster_lvl2,
  cluster_lvl3 = NULL)

Arguments

eq_lvl1	Regression equation for level-1 effects.
eq_lvl2	List of regression equations for level-2 effects.
eq_lvl3	List of regression equations for level-3 effects.
cluster_lvl2	Name of grouping variable that defines level-2 clusters.
cluster_lvl3	Name of grouping variable that defines level-3 clusters.

Value

A list containing (1) a mixed-effects equation, (2) a full level-1 equation, and (3) a level-1 equation for random effects only.

Examples

## Level-1 equation:
eq_lvl1 <- "outcome ~ x1_lvl1 + x2_lvl1"

## Level-2 equations:
eq_lvl2 <- list("outcome ~ x1_lvl2 + x2_lvl2",
                "x1_lvl1 ~ x1_lvl2 + x2_lvl2",
                "x2_lvl1 ~ 1")

## Level-3 equations:
eq_lvl3 <- list("outcome ~ x1_lvl3",
                "x1_lvl2 ~ x1_lvl3",
                "x2_lvl2 ~ 1")

## Clustering variables:
cluster_lvl2 <- "cluster2"
cluster_lvl3 <- "cluster3"

## Two-level model with no level-2 explanatory variables:
create_regeq(eq_lvl1 = eq_lvl1, eq_lvl2 = NULL,
             cluster_lvl2 = cluster_lvl2)

## Two-level model with level-2 explanatory variables:
create_regeq(eq_lvl1 = eq_lvl1, eq_lvl2 = eq_lvl2,
             cluster_lvl2 = cluster_lvl2)

## Three-level model:
create_regeq(eq_lvl1 = eq_lvl1, eq_lvl2 = eq_lvl2, eq_lvl3 = eq_lvl3,
             cluster_lvl2 = cluster_lvl2, cluster_lvl3 = cluster_lvl3)

jadahlke/hlmer documentation built on May 31, 2019, 8:43 a.m.