Time: June 4, 2015 at 9am to June 6, 2015 at 5pm
Location: Temple University Center City
Street: 1515 Market Street
Website or Map: http://www.statisticalhorizon…
Event Type: training, seminar
Organized By: Statistical Horizons LLC
Latest Activity: Feb 12, 2015
Taught by Gregory R. Hancock, Ph.D.
Structural equation modeling (SEM) is a versatile analytical framework for estimating and assessing models that describe relations among both measured and latent variables. Common examples include measured variable path models, confirmatory factor models, and latent variable path models. These models subsume methods based on the traditional general linear model such as multiple regression and analysis of variance.
This seminar goes beyond introductory SEM to cover more advanced methods that enable researchers to address their current modeling questions more effectively and also to focus on entirely new research questions. After a review of SEM basics, we will cover real data challenges (e.g., missing data, nonnormality, complex samples, categorical data), mean structure models for measured and latent variables, models for latent variable interactions, latent growth curve models, and power analysis in SEM. The style of instruction is designed for participants with a variety of content backgrounds. Examples will be presented using both Mplus and the SIMPLIS language within LISREL.
This is a hands-on course with lots of in-class practice exercises. Participants are strongly encouraged to bring their own laptop PCs to use either Mplus (for those students who already have the software) or the free student version of LISREL (which participants will be instructed on how to download prior to the course). Participants may use other SEM packages as well, although explicit in-class support will not be provided.