minvariance

The goal of minvariance is to aid the understanding of longitudinal measurement invariance and see which parts of this method could be automated in R. The vignettes on the website of this package aim to replicate some examples found online. If you have any questions, ideas, or comments please get in touch.

Installation

You can install the released version of minvariance from GitHub with:

# install.packages("devtools")
devtools::install_github("milanwiedemann/minvariance")

library(tidyverse)
#> ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.0 ──
#> ✓ ggplot2 3.3.3     ✓ purrr   0.3.4
#> ✓ tibble  3.1.0     ✓ dplyr   1.0.5
#> ✓ tidyr   1.1.3     ✓ stringr 1.4.0
#> ✓ readr   1.4.0     ✓ forcats 0.5.1
#> ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
#> x dplyr::filter() masks stats::filter()
#> x dplyr::lag()    masks stats::lag()
library(minvariance)

# Specify list of variables 
# Generate lavaan syntax
long_minvariance_syntax(var_list = list(t1 = c("i1_t1", "i2_t1", "i3_t1"), 
                                        t2 = c("i1_t2", "i1_t2", "i1_t2")), 
                        model = "configural") %>% 
  cat()
#> Configural Invariance Model (Pattern Invariance)
#> #### CONFIGURAL INVARIANCE MODEL ####
#> # Specify Latent Factors ----
#> eta1 =~ NA * i1_t1 + lambda1 * i1_t1 + i2_t1 + i3_t1
#> eta2 =~ NA * i1_t2 + lambda1 * i1_t2 + i1_t2 + i1_t2
#> # Specify Latent Variable Means ----
#> eta1 ~ 0 * 1 
#> eta2 ~ 1 
#> # Specify Latent Variable Variances ----
#> eta1 ~~ 1 * eta1
#> eta2 ~~ eta2
#> # Specify Latent Variable Covariances ----
#> eta1 ~~ eta2
#> # Specify Observed Variable Intercepts ----
#> i1_t1 ~ tau1 * 1 
#> i2_t1 ~ 1 
#> i3_t1 ~ 1 
#> i1_t2 ~ tau1 * 1 
#> i1_t2 ~ 1 
#> i1_t2 ~ 1 
#> # Specify Unique Variances ----
#> i1_t1 ~~ i1_t1 
#> i2_t1 ~~ i2_t1 
#> i3_t1 ~~ i3_t1 
#> i1_t2 ~~ i1_t2 
#> i1_t2 ~~ i1_t2 
#> i1_t2 ~~ i1_t2 
#> # Specify Unique Covariances ----
#> i1_t1 ~~ i1_t2
#> i2_t1 ~~ i1_t2
#> i3_t1 ~~ i1_t2

Here’s another example with different variable names

# Generate lavaan syntax
long_minvariance_syntax(var_list = list(t1 = c("a1", "b1", "c1"), 
                                        t2 = c("a2", "b2", "c2")), 
                        model = "strong") %>% 
  cat()
#> Strong Invariance Model (Scalar Invariance, Intercept Invariance)
#> #### STRONG INVARIANCE MODEL ####
#> # Specify Latent Factors ----
#> eta1 =~ NA * a1 + lambda1 * a1 + lambda2 * b1 + lambda3 * c1
#> eta2 =~ NA * a2 + lambda1 * a2 + lambda2 * b2 + lambda3 * c2
#> # Specify Latent Variable Means ----
#> eta1 ~ 0 * 1 
#> eta2 ~ 1 
#> # Specify Latent Variable Variances ----
#> eta1 ~~ 1 * eta1
#> eta2 ~~ eta2
#> # Specify Latent Variable Covariances ----
#> eta1 ~~ eta2
#> # Specify Observed Variable Intercepts ----
#> a1 ~ tau1 * 1 
#> b1 ~ tau2 * 1 
#> c1 ~ tau3 * 1 
#> a2 ~ tau1 * 1 
#> b2 ~ tau2 * 1 
#> c2 ~ tau3 * 1 
#> # Specify Unique Variances ----
#> a1 ~~ a1 
#> b1 ~~ b1 
#> c1 ~~ c1 
#> a2 ~~ a2 
#> b2 ~~ b2 
#> c2 ~~ c2 
#> # Specify Unique Covariances ----
#> a1 ~~ a2
#> b1 ~~ b2
#> c1 ~~ c2

Resources and related work

R functions to test for longitudinal factorial/measurement invariance

• Longitudinal Factor Analysis - Measurement Invariance by Nilam Ram
• Measurement Matters Resource List by Eiko Fried and Jessica Flake
• Fried et al (2016). Measuring depression over time . . . Or not? Lack of unidimensionality and longitudinal measurement invariance in four common rating scales of depression. Psychological Assessment. and Supplemental Materials with Mplus files
• Grimm (2017) - lavaan Code for Chapter 14 - Modeling Change with Latent Variables Measured by Continuous Indicators by Julie Wood
• MIE - Measurement invariance explorer (R package) by Maksim Rudnev
• “A Tutorial in Longitudinal Measurement Invariance Using Lavaan” by Robin Curtis, Sean Mackinnon, Roisin O’Connor
• measEq.syntax {semTools}
• On dimensionality, measurement invariance and suitabilityof sum scores for the PHQ-9 and the GAD-7 with Mplus Code using mplusautomation
• lavaan group