Calendar of events

Hosts: Dr Suzanne Aussems and Dr Chiara Gambi

Speaker: Dr Sam Cacace from the School of Psychology and Clinical Language Sciences, University of Reading

Abstract

Psychologists and other social scientists often deal in abstract concepts, like emotions, moral judgement, depression, trauma, and stress. Self-report scales, biofeedback, participant observation, and other commonly used methods provide means to access these valuable parts of human experience, but can often be unreliable, exhibit poor validity, and/or fail to fully address the underlying construct. Fortunately, structural equation modelling (SEM) offers ways to evaluate relationships between constructs that are not directly observable (i.e., latent variables), through a two-step process.

In this interactive workshop, you will learn the foundations of SEM and the types of questions it can answer, including evaluating measurement models through confirmatory factor analysis, and structural models. We will work through SEM examples, ensuring model identification, power analysis and sample size checking, measurement reliability, and guidance on how to start thinking about and using latent variables in your own research. All coding will be tailored to R, but will be accessible to those who choose other statistical packages or are relatively new to the software, with additional resources provided to explore the techniques.

Time will be allocated at the end of the workshop for you attempt to code your own SEM and receive personalized assistance.

Preparation: This workshop will be run in R and Rstudio so please bring a laptop with these installed. You can also pre-install some of the packages that the workshop will be using, by running these two lines of code in R:

install.packages(c("haven", "dplyr", "lavaan", "semPower", "semTools", "semptools"))
devtools::install_github("melissagwolf/dynamic")

There will be a social lunch around 1pm, where we will go somewhere on campus to eat. The lunch will be covered, so please sign up here: Introduction to Structural Equation Modelling