SEM Using AMOS | Data Mining | Analysis Of A Moment Structure | Big Data Analytics | Structural Equation Modelling

1
SEM Using AMOS
2
Statswork
www.statswork.com
Statistical Consulting Company
3
What is SEM Structural Equation Modelling (SEM)
is a widely used technique in statistics to
primarily study relationships based on
structures. It encompasses various models
involving mathematics, statistical procedures
etc. This technique is known to be extremely
effective when it comes to measuring latent
constructs.
4

SEM Using AMOS
5
Benefits Of SEM
Here are some of the significant benefits of
using Structural Equation Modelling as a
technique If you are a researcher looking to
expand your scope, using SEM would be the ideal
choice for the assumptions which brings a lot of
clarity and they are testable too. It enables
survey sampling analyses. Coefficients, means and
variances from different subjects can be compared
at once. You can use models that are not standard
including databases containing data which is not
enough and incorrectly distributed.
6
Functioning of SEM As a researcher, you ought to
begin by choosing a model. And, you have to
collect data only after figuring out how to
evaluate constructs. Finally, you supply the SEM
software with sufficient amount of data. The
software then fits the data to the chosen model
and generates the outcome. The outcome would
usually include estimates and overall model fit
figures.
7
SEM And AMOS
AMOS
SEM
SEM as a technique is largely dependent on this
statistical software called AMOS (Analysis of
Moment Structures). It produces tabular results
similar to the ones, one can see in SPSS,
considering it is an added module of the same.
Structural Equation Modelling (SEM) is a widely
used technique in statistics to primarily study
relationships based on structures. It encompasses
various models involving mathematics, statistical
procedures etc.
8
SEM Using AMOS
9
Use Big Images To Show Your Ideas
10

• Methods used by AMOS
• UnWeighted Least Squares It eliminates residual
  errors in order to access the conditional mean.
• Generalised Least Squares It estimates the
  coefficients in a linear regression model if some
  correlation exists amongst the residuals.
• Generalised Least Squares It estimates the
  coefficients in a linear regression model if some
  correlation exists amongst the residuals.

11

• Model Construction
• Data Input You will need to enter your data for
  the purpose of SEM Analysis. Choose a name for
  your file and record your data in AMOS.
• Icons Go with Rectangle and Circle icons for
  observed and unobserved variables respectively.
• Establishing Relationships Draw an arrow to
  denote the relationship between observed and
  unobserved variables.
• Covariance Choose a double-headed arrow to
  denote the covariance amongst variables.
• Error Term The icon denoting the same is
  situated next to the unobserved variable icon.
  The Error Term icon is present to chart the
  latent variable.

12

• Text Results in AMOS
• While graphic window will only show you some part
  of the data including standardized and
  unstandardized regressions, text output will
  reveal the results in its entirety.
• Number of Variables The number of observed and
  unobserved variables used in the process of SEM
  analysis will be revealed.
• Data normality It is important that the data
  used in SEM analysis is normally distributed. The
  text output of AMOS will help us gauge the
  normality of data.
• Impact of Path Analysis Modification Index
  results tell us how impactful the path drawn by
  you can be, if the index is high, it is a sign
  for you to draw more paths.

13
Thanks!
Any questions?
You can find me at info_at_statswork.com www.statswo
rk.com