Multilevel Modeling

Course Overview and Workflow Recap

Spring 2026 | CLAS | PSYC 894
Jeffrey M. Girard | Lecture 01b

Roadmap

Syllabus Review
Course Modules
Workflow Recap

Syllabus Review

Instructor Information

Jeffrey Girard, PhD
affcom.ku.edu/girard
jmgirard@ku.edu
Fraser Hall 454

Background

Associate Professor, Department of Psychology
Brain, Behavior, and Quantitative Science
Clinical Psychological Science
Kansas Data Science Consortium

Research Areas

Affective Computing
Mental and Social Health
Applied Statistics and Machine Learning
Data Science and Software Engineering

Learning Outcomes

Learn fundamental concepts underlying multilevel modeling
Learn practical skills for implementing such models in R
Instill a sense of mastery, curiosity, and fun

What you will learn here

When and why MLM is needed
Data prep and model building
Diagnostics and interpretation
Exposure to advanced topics
Exposure to alternatives

What you won’t learn here

Mastery of longitudinal models
Mastery of nested and crossed models
Mastery of Bayesian estimation
Mastery of Extensions (GLMM, GAMM)
Mastery of Alternatives (WLS, GEE)

Course Materials

Textbooks

Introduction to Multilevel Modeling [S&F]*
Shaw & Flake (2021, 1E, www.learn-mlms.com, free)
Multilevel analysis: Techniques and applications [HMV]
Hox, Moerbeek, & van de Schoot (2017, 3E, www.taylorfrancis.com, paid)
R for Data Science: Import, tidy, transform, visualize, and model data [WCG]
Wickham, Çetinkaya-Rundel, & Grolemund (2023, 2E, r4ds.hadley.mz, free)

R Packages

Frequentist Model Fitting: lme4*, glmmTMB
Bayesian Model Fitting: rstanarm, brms
Utilities and Visualization: easystats*, ggeffects*

Grading Policies

Graded Components

66% Lab Reports (6% each \(\times\) 11) - Practice skills and concepts from lecture
22% Reflections (2% each \(\times\) 11) - Summarize and consolidate learning from reading
12% Final Project (12% each \(\times\) 1) - Apply skills to real-world data

Grading Scale

Letter	A	A−	B+	B	B−	C+	C	C−	D+	D	D−
G. Points	4.0	3.7	3.3	3.0	2.7	2.3	2.0	1.7	1.3	1.0	0.7
Threshold	93%	90%	87%	83%	80%	77%	73%	70%	67%	63%	60%

I don’t round or curve grades, but will give points back if a question was poorly designed.

Course Policies

Attendance

In-person attendance is strongly encouraged but not required
Video recordings will be attempted but cannot be guaranteed

Late Submissions

Detailed on each rubric, partial credit is possible for late submissions
However, no late submissions will be accepted after Stop Day (May 9)

Academic Misconduct

Acceptable: peers, tutors, technology used to explain or troubleshoot
Unacceptable: peers, tutors, technology used to complete assignments
You must use the skills and tools taught in class to solve the lab reports

Course Materials

Canvas

Has assignments, grades, videos, and textbook PDFs
https://canvas.ku.edu/courses/182539

Website

Has topics, ebook links, datasets, and slideshows
https://jmgirard.github.io/psyc894
Note this may be renamed psyc892 next year

Course Modules

Unit A Overview

Week 01: Course Overview and Recap of R
- No deadlines
Week 02: Recap of Linear Modeling
- Lab Report only
Week 03: Clustering Issues and Approaches
- Reflection and Lab Report
Week 04: Data Prep and Estimation
- Reflection and Lab Report

Unit B Overview

Week 05: Random/Varying Intercepts
- Reflection and Lab Report
Week 06: Random/Varying Slopes
- Reflection and Lab Report
Week 07: Cross-level Interactions
- Reflection and Lab Report
Week 08: Translation Activity
- Reflection only

Unit C Overview

Week 09: Spring Break
- No deadlines
Week 10: Model Comparison
- Reflection and Lab Report
Week 11: Longitudinal Analyses
- Reflection and Lab Report
Week 12: Diagnostics and Reporting
- Reflection and Lab Report

Unit D Overview

Week 13: Troubleshooting and Bayes
- Reflection and Lab Report
Week 14: Advanced Previews
- Reflection and Lab Report
Week 15: Project Work
- No deadlines
Week 16: Project Work
- Project Report

Workflow Recap

The R Ecosystem

Think of your computer as the engine of a car
- It provides raw power for computation

The R language is like the controls for the car
- It lets you apply and direct that power

RStudio is like a fancy dashboard for the car
- It adds extra information and convenience

An R package is like an add-on for the car
- It adds new features and capabilities

Communicating with R

The Console is like a chat window with R
- You send a command to R and get a response
- Neither side of the conversation is saved

An R Script is like an email thread with R
- You send many commands to R all at once
- Only your side of the conversation is saved

A Quarto Document is like a scrapbook with R
- You can combine code and formatted text
- Both sides of the conversation are saved

Projects

File Management

Projects are special folders on your computer
- They contain all files related to a task
- They keep everything together and organized

Projects make it easy to find and use your files
- No need to specify long, annoying file paths
- No need to worry about working directories

Projects make it easy to switch between tasks
- They will remember exactly where you left off
- You can even open multiple projects at once

Create a new Project

Open the “File” menu in RStudio
Select the “New Project…” option
Select the “New Directory” option
Select the “New Project” option
Name the directory “PSYC 894” or “MLM” (or whatever)
Browse to where to create your Project folder

Create a new File

Explore the Files tab in the Extras pane
Create a New File (e.g., a script) as an example
RStudio will automatically create it in your project folder
Add some text to the example file (e.g., “# Hello World”)
Save the file and note that it defaults to your project folder
Close the script with the “x” icon
Reopen the script from the Files tab

Close and Open Projects

Open the “File” menu in RStudio
Select the “Close Project” option
Notice that your work is now gone
Open the “File” menu in RStudio
Select the “Open Project” option
Browse to your project folder
Open the .Rproj file (e.g., PSYC 399.Rproj)
Notice that your work is now back!

Quarto

Quarto is a scientific/technical publishing system

Create dynamic content (e.g., code, output, and markdown) with R, Python, Julia, and Observable

Articles, reports, slideshows, websites, blogs, and books in HTML, PDF, MS Word, and ePub formats

Include equations, citations, crossrefs, figure panels, callouts, advanced layouts, etc.

Create an Quarto Document

Open the “File” menu in RStudio
Select the “New File” option
Select the “Quarto Document…” option
Keep the defaults (HTML) and hit “Create”
Open the “File” menu
Select the “Save” option
Note that it defaults to the project folder
Give it a name like “Lecture 02c” (or whatever)
Note that the file extension is .qmd

Render the boilerplate content

To see how it works, read the boilerplate content
Now click the “Render” button to create the output file
A preview of the output will appear in the Viewer tab
In the File tab, we see a new .html file was created
We can share this .html file with others
It includes all the formatted text, code, and R results

Remove the boilerplate content

The top part of the document is called the “Header”
You can change the title but keep it in quotes
Highlight and delete everything below the header
In the future, you can click “Create Empty Document”
Add embed-resources: true to the Header for assignments!

Add an R Chunk

Option 1: Click the green box with C and a plus sign icon (top-right)
Option 2: Ctrl+Alt+I (Win) or Cmd+Option+I (Mac)
Option 3: Type out ```{r} then new line and ``` yourself

Adding code to the chunk

Anything you put inside the “fences” (i.e., ```) will be treated as R code
The chunk operates like a mini console
Try doing some calculations in the chunk and hit the green arrow
The answer appears right below the chunk!
If we save and render the document, it appears in the output too
We can add options to the chunk but we will return to this later

Markdown

Markdown is a simple text-to-HTML conversion language used by Quarto, GitHub, Obsidian, etc.

It will allow us to add formatted text, images, links, lists, etc. to our Quarto documents

Quarto will treat all text that is not in a code chunk as markdown

Formatting

Markdown	Output
`italics, bold, *bold italics*`	italics, bold, *bold italics*
`superscript^2^ / subscript~2~`	superscript² / subscript₂
`~~strikethrough~~`	~~strikethrough~~
`verbatim code`	`verbatim code`

Headings

Markdown	Output
`# Heading 1`	Heading 1
`## Heading 2`	Heading 2
`### Heading 3`	Heading 3
`#### Heading 4`	Heading 4

R Refresher

Mathematical Operations: + - * / ^ ( )
Assignment and Naming: <-
Storing Text in Strings: "text data"
Using Built-in Functions: sqrt(64)
Creating and Using Vectors: c(1, 2, 3)
Installing New Packages: install.packages("lme4")
Importing Data from Files: df <- read.csv("file.csv")

Multilevel Modeling

Roadmap

Syllabus Review

Instructor Information

Learning Outcomes

Course Materials

Textbooks

R Packages

Grading Policies

Graded Components

Grading Scale

Course Policies

Attendance

Late Submissions

Academic Misconduct

Course Materials

Canvas

Website

Course Modules

Unit A Overview

Unit B Overview

Unit C Overview

Unit D Overview

Workflow Recap

The R Ecosystem

Communicating with R

Projects

File Management

Create a new Project

Create a new File

Close and Open Projects

Quarto

Create an Quarto Document

Render the boilerplate content

Remove the boilerplate content

Add an R Chunk

Adding code to the chunk

Markdown

Markdown

Formatting

Headings

Heading 1

Heading 2

Heading 3

Heading 4

R Refresher