---
title: "RICLPM Workshop"
output: html_document
editor_options: 
  chunk_output_type: console
---

Welcome to the workshop titled "Decomposing Within & Between Person Effects in Longitudinal Data with SEM in R"! My name is Dustin Haraden and I will be somewhat of a guide throughout this process. Let's hope that we can work together to get you comfortable enough to start asking some more interesting questions! 

I am an Assistant Professor in Clinical Psychology at Rochester Institute of Technology. My research interests focus on examining risk for depression in youth with a focus on sleep, circadian rhythms and pubertal development. I also have an interest in methodology, statistics and open science. 

Please feel free to reach out for any reason! dxhgsh@rit.edu or you can follow me on blue sky @dustinharaden.bsky.social

Let's get started! 

```{r setup, include=FALSE}
# for SEM
library(lavaan)
# for any manipulation
library(dplyr)
# some data processing (may not use)
library(tidyr)
# visualizations
library(ggplot2)
# SEM Visualizations (sometimes)
library(semPlot)

dat <- read.table("https://raw.githubusercontent.com/JeroenDMulder/RI-CLPM/refs/heads/master/data/RICLPM.dat", 
                  col.names = c(
                    "x1", "x2", "x3", "x4", "x5", 
                    "y1", "y2", "y3", "y4", "y5")
) 

```

Loading in the data above that has accompanied the article from Mulder, J. D., & Hamaker, E. L. (2021). Three extensions of the random intercept cross-lagged panel model. Structural Equation Modeling: A Multidisciplinary Journal, 28(4), 638-648. 

This will be the main source of data for running the models for the current workshop. If there is the need for more, that will be shared with you. 

```{r}
clpm_syntax <- '
    #regressions
    x2 ~ a*x1
    x3 ~ a*x2
    x4 ~ a*x3
    
    y2 ~ b*y1
    y3 ~ b*y2
    y4 ~ b*y3
    
    # cross-paths
    
    y2 ~ c*x1
    y3 ~ c*x2
    y4 ~ c*x3
    
    x2 ~ d*y1
    x3 ~ d*y2
    x4 ~ d*y3
    
    # covariances
    x1 ~~ y1
    x2 ~~ y2
    x3 ~~ y3
    x4 ~~ y4
'


#Fit the model 
clpm.fit <- sem(clpm_syntax, data = dat, missing = 'ML')
summary(clpm.fit)

semPaths(clpm.fit, layout = "tree2")
```


 Now we will introduce the random intercept cross lagged panel model 
 
```{r}
riclpm_syntax <- '
    # Random Intercepts
    RIx =~ 1*x1 + 1*x2 + 1*x3 + 1*x4
    RIy =~ 1*y1 + 1*y2 + 1*y3 + 1*y4
    
      #RI variance Covariance
      RIx ~~ RIy
      RIx ~~ RIx
      RIy ~~ RIy
      
    # Within Person
    wx1 =~ 1*x1
    wx2 =~ 1*x2
    wx3 =~ 1*x3
    wx4 =~ 1*x4
    wy1 =~ 1*y1
    wy2 =~ 1*y2
    wy3 =~ 1*y3
    wy4 =~ 1*y4


    # within person regressions
    wx2 ~ a*wx1
    wx3 ~ a*wx2
    wx4 ~ a*wx3
    
    wy2 ~ b*wy1
    wy3 ~ b*wy2
    wy4 ~ b*wy3
    
    # cross-paths
    
    wy2 ~ c*wx1
    wy3 ~ c*wx2
    wy4 ~ c*wx3
    
    wx2 ~ d*wy1
    wx3 ~ d*wy2
    wx4 ~ d*wy3
    
    # covariances
    wx1 ~~ wy1
    wx2 ~~ wy2
    wx3 ~~ wy3
    wx4 ~~ wy4
    
    # varaince
    wx1 ~~ wx1
    wx2 ~~ wx2
    wx3 ~~ wx3
    wx4 ~~ wx4
    wy1 ~~ wy1
    wy2 ~~ wy2
    wy3 ~~ wy3
    wy4 ~~ wy4
'

riclpm.fit <- lavaan(riclpm_syntax, data = dat)
summary(riclpm.fit, fit.measures = TRUE)

parameters::model_parameters(riclpm.fit)
```