Introduction to Multilevel Measurement Models
SMIP Summer School 2025: Lecture 01
Workshop Introduction
This is an ambitious workshop:
- Measurement models (Item Response Theory)
- Multilevel models for clustered data
- Bayesian statistics
- Multilevel measurement models
Each could be its own workshop
Workshop Materials
Set of slides on topics pulled from other course materials
- In case additional detail on any topic is needed
Full set of analysis files in R and Stan
- In case you would like to work ahead
Workshop Schedule
- Mornings: 9:00-12:30 (Including Friday)
- Afternoons:
- Tuesday: 14:00-17:15 (break at 16:00)
- Wednesday: 15:15-17:45 (break at 16:45)
- Thursday: 14:00-15:45 (no break)
Lecture and Example Time Plan
Last hour of each day is open
Your choice:
- Questions
- Example practice
- Work with your data
- Leave early
Workshop Sections
- Introduction to Measurement Models, Bayesian Statistics, and Stan
- Introduction to Multilevel Models
- Introduction to Multilevel Measurement Models
Running Example Data
Data come from a secondary school (10th grade) end-of-grade mathematics assessment given around the year 2006 in a midwestern state in the USA
- Sample of 50 students from 62 schools
- Sample of 10 mathematics items from the assessment
- Items are scored correct/incorrect (
score1
-score10
)
- The sum of all 10 items is included (
sumScore
)
- Other variables:
studentID
: Student ID number (created for this example)
schoolID
: School ID number (created for this example)
frlunch
: Free/reduced lunch status (1 = free/reduced lunch, 0 = not free/reduced lunch; indirect indicator of student socioeconomic status)
Distribution of Sum Scores
Distribution of Item Difficulties
Workshop Lectures
- Lecture 01: Introduction to Multilevel Measurement Models Workshop
- Lecture 02: Introduction to Measurement Models
- Lecture 03: Introduction to Bayesian Statistics
- Lecture 04: Introduction to MCMC and Stan
- Lecture 05: Bayesian IRT Models in Stan
- Lecture 06: Introduction to Multilevel Models
- Lecture 07: Bayesian Multilevel Models in Stan
- Lecture 08: Bayesian Multilevel Measurement Models in Stan