HKSQ and TQM co-organized webinar
on “Using Structural Equation Modeling (SEM) to Determine the Associations
Among Factors that Drive Performance” on 9th Dec 2022. CityU EngD
Society was a supporting organization. This webinar gave a brief overview of
what SEM is, how it differs from common statistical models such as linear
regression, and when it may be more useful than other statistical models. Before
the webinar, Dr. Jane Wong (Chairman, HKSQ) introduced Hong Kong Society for
Quality and the guest speaker – Dr. Frank Reichert (Assistant Professor of
Interprofessional Education, Faculty of Education, The University of Hong Kong,
HK SAR).
In the beginning, Dr. Frank Reichert introduced
what is structural equation modeling (SEM) that can model unobserved, latent
variables. It estimates a set of regressions simultaneously and examine the fit
between theoretical model and empirical data.
Then Dr. Reichert mentioned the measurement in
SEM based on theoretical assumptions (or hypotheses). They often employed
theoretical constructs that cannot be observed such as aggression, motivation,
intelligence, self-efficacy and project performance, etc.
The construct means latent variable that cannot
be measured directly. After that he briefed when to use SEM. It is dealing with
latent variables that have clear hypotheses and would like to test the
relationship between multiple variables.
And then different models are introduced
including Path Models, Confirmatory Factor Analysis and Confirmatory Composite
Analysis, as well as, Structural Regression Model, etc.
Finally, he introduced different SEM software
for application.
Some literatures are suggested for further
study.
Lastly, we had Q&A session to discuss the
application of SEM in real case.
沒有留言:
發佈留言