This is the home page for Psychology 454: Psychological Measurement: Latent Variable Modeling.
Much of the current literature in psychology and related disciplines includes variables which are not directly observable but rather are inferred from the observed variables. That is, our observable measures are thought to reflect some underlying latent process or processes. Our theories tend to be formulated in terms of these latent processes. The problem then is how to measure or model these unobservable processes. This is the challenge of this course.
The objectives of this course are
- To understand the fundamental concepts in latent variable modeling in order to make you a better con- sumer and producer of latent variable models in your research. This includes a number of latent variable techniques, including, but not limited to Exploratory Factor Analysis (EFA), Confirmatory Factor Analysis (CFA), Structural Equation Modeling (SEM) and Item Response Theory (IRT). By necessity, it also includes fundamental issues in reliability.
- Thus, the goal of this course is to help you understand how to evaluate the quality of models when applied to data by understanding various sources of variability as measured by goodness of fit tests. The emphasis will be upon model comparison rather than model confirmation.
- A further goal is to learn how to apply these concepts to real data sets using a variety of standard statistical packages (in particular in R (R Core Team, 2020) using the lavaan package (Rosseel, 2012)). Some examples will be given from other software (e.g., EQS, Lisrel, Prelis, Amos, and MPlus).
The course syllabus is available here .
Homework assignments for the course will appear here:
- Course notes for the first week
- A review of matrix algebra
- Go to the syllabus for detailed list of readings and handouts.
The R statistical system is particularly useful for modeling latent variables. Almost all of the examples will make use of R. To help you learn R, there are a a number of tutorials and short courses for using R. These are useful background for this course and are here:
- A four hour course for the APS
- A two hour course for the APA (part 1)
- A two hour course for the APA (part 2)
For more even more practice in R, see Psychology 350: Special Topics: Using R for Psychological Research .
This page is part of the Personality Project.
If you want to help us develop our understanding of personality, please take our test at SAPA Project. The SAPA project is a collaborative data collection tool for assessing psychological constructs across multiple dimensions of personality. These dimensions currently include temperament, cognitive abilities, and interests (TAI) because research suggests that these three dimensions cover a large percentage of the variability among people without too much redundacy. Other broad dimensions like motivation and character may be included later depending on how much incremental benefit they provide in terms of predicting outcomes.
After taking the survey you will get feedback about your personality on each of these dimensions relative to other people of your age and gender. The feedback will be customized to your responses using the scores of approximately 1,000,000 people who have participated so far.
The test is entirely free (absolutely no strings attached) and completely anonymous (no cookies, no tracking, nothing). And, you will have our gratitude -- this research is not possible without the help of people like you!