Introduction to Structural Equation Modelling (with MPlus 8.1 & LISREL 9.3)
Speaker: Mr Jerome Oko: Phd Candidate, School of Education, University of Adelaide.
Scope/Abstract: Structural Equation Modelling (SEM), first introduced in the early 70s, is a statistical methodology used widely by researchers in the human sciences. It brings together techniques used in quantitative methods, psychometrics, econometrics and epidemiology. The concept of latent (hidden) construct is central in the human sciences, and is modelled through observed manifest variables by partitioning out the residuals or error variances.
SEM is a multivariate statistical analysis technique used to analyse structural relationships between measured variables and latent constructs, and is the combination of factor analysis and multiple regression analysis.SEM is used to estimate multiple and interrelated dependence in a single analysis.
This workshop provides an overview of SEM, and introduces the two models used in SEM: measurement (specifies how measured variables ‘combine’ to represent a construct or theory) and structural (tests proposed causal relationships and interactions) models. Complex relationships between manifest variables and/or latent constructs can be tested in path-models which are not possible to specify in the multiple regression. Confirmatory factor analysis (CFA) and simple path models are examined in the workshop. The use of both MPlus (Version 8.1) and LISREL (Version 9.3) will be demonstrated.
Biography: Don Bosco Technological Institute (Papua New Guinea) Academic Head (2012-2015) Mathematics Tutor (2007-2009)
2010-2011 Master of Education (Mathematics & Technology) Don Bosco Technological Institute an Affiliate of Divine Word University (Papua New Guinea)
2003-2006 Bachelor of Education-Technical Degree
Requirements: Participants need to understand the basics of multiple regression analysis, and have access to a laptop for use during the workshop.
Software: Please download and install the following programs:
MPlus Demo Version 8.1 (https://www.statmodel.com/demo.shtml), and
LISREL 9.3 Student Version (http://www.ssicentral.com/lisrel/student.html).