sem_book's Introduction

An Introduction to Structural Equation Modeling in R

Structural equation modeling is among the fastest growing statistical techniques in the natural sciences, thanks in large part to new advances and software packages that make it broadly applicable and easy to use.

This book is meant to be an approachable and open-source guide to the theory, math, and application of SEM. It integrates code for the R software for statistical computing from popular packages such as lavaan and piecewiseSEM. Each chapter ends with worked examples from the published literature.

Moreover, as the author of the piecewiseSEM package, this format allows me to document newly-deployed functionality in the package, such as the addition of categorical variables, multigroup analysis and composite variables, new forms of coefficient standardization, and updates to model R²s.

Check back often, as this book is a "living resource:" as new functionality is added and bugs uncovered and fixed, they will be described in detail here (with worked examples where possible).

In some ways, this is my grandest experiment: happy reading!

sem_book's People

Stargazers

Watchers

sem_book's Issues

LLchisq() not an available function

Hi! In Chapter 3.4 you go over the Log-Likelihood chi-squared test and make reference to a function LLchisq() , which I presume should be in your piecewiseSEM package, but I don't see it in there (or in any other R package) and am not able to run it. Very nice book by the way, thanks for this resource!

How to cite the SEM book?

Are there plans in the works to get a DOI for the book or the repository? Do you know of any papers citing the work?

I find myself using this book quite a bit and want to be sure you're properly credited!

Categorical variables example

The Spartina genotype SEM example has below.C as the response and belowCN as the predictor

Typo in section 3.6

Hi there is a very minor typo in the 4th line of the 3rd code chunk in section 3.6
gam should be glm (in order to get a pval of 0.647)
Great book and package by the way!
Claire

shipley_psem3 <- psem(
gam(x2 ~ s(x1), data = dat2, family = gaussian),
glm(x3 ~ x2, data = dat2, family = poisson),
gam(x4 ~ x2, data = dat2, family = poisson), #should be glm
glm(x5 ~ x3 + x4, data = dat2, family = binomial)
)

Recommend Projects