Summary
Contents
Subject index
This student orientated guide to structural equation modeling promotes theoretical understanding and inspires students with the confidence to successfully apply SEM. Assuming no previous experience, and a minimum of mathematical knowledge, this is an invaluable companion for students taking introductory SEM courses in any discipline. Niels Blunch shines a light on each step of the structural equation modeling process, providing a detailed introduction to SPSS and EQS with a focus on EQS’ excellent graphical interface. He also sets out best practice for data entry and programming, and uses real life data to show how SEM is applied in research. The book includes: • Learning objectives, key concepts and questions for further discussion in each chapter. • Helpful diagrams and screenshots to expand on concepts covered in the texts. • A wide variety of examples from multiple disciplines and real world contexts. • Exercises for each chapter on an accompanying companion website. • A detailed glossary. Clear, engaging and built around key software, this is an ideal introduction for anyone new to SEM.
Incomplete and Non-normal Data
Incomplete and Non-normal Data
Missing data are more the rule than the exception in empirical research, and several solutions to the problem have been suggested. As will be argued in the opening paragraphs of this chapter, the most widespread ones are not wholly satisfactory.
Then you will learn about two more satisfying methods, the first of which is full information maximum likelihood estimation that makes it possible to estimate a model using all data at hand even if some data are missing. However, it turns out that even this technique is not without its drawbacks – the most serious being that you easily run out of degrees of freedom, because it is necessary to estimate means and intercepts.
Another possibility is to use multiple imputation, a ...
- Loading...
