0% found this document useful (0 votes)
60 views29 pages

Workshop On SEM Intro NRMS

The document describes a workshop on structural equation modeling (SEM) using AMOS software. The workshop was scheduled for June 27-28, 2016 and was to be led by Associate Professor Dr. Nur Riza Mohd Suradi. The workshop outline included introductions to SEM and AMOS, as well as sessions on path analysis, confirmatory factor analysis, structural equation modeling, and analyzing, interpreting and reporting SEM results. Example research topics that SEM can be applied to were also provided such as relationships between measured and unmeasured variables.

Uploaded by

nur riza
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
60 views29 pages

Workshop On SEM Intro NRMS

The document describes a workshop on structural equation modeling (SEM) using AMOS software. The workshop was scheduled for June 27-28, 2016 and was to be led by Associate Professor Dr. Nur Riza Mohd Suradi. The workshop outline included introductions to SEM and AMOS, as well as sessions on path analysis, confirmatory factor analysis, structural equation modeling, and analyzing, interpreting and reporting SEM results. Example research topics that SEM can be applied to were also provided such as relationships between measured and unmeasured variables.

Uploaded by

nur riza
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 29

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS

WORKSHOP ON
SEM USING AMOS

ASSOC. PROF. DR. NUR RIZA MOHD SURADI


Centre for Modelling and Data Analysis (DELTA)
School of Mathematical Sciences, FST, UKM
[email protected]

Name:
Workshop Outline
Introduction to SEM
Introduction to AMOS
Path Analysis
Confirmatory Factor Analysis
Structural Equation Modeling
Analysis, Interpreting and Reporting

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 2


MODULE 1
Introduction to SEM:
Why and When
Essential Concepts & Terminologies

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS

3
RESEARCH ISSUES
• Often in research, we study the relationships between variables, or the
influences/ effects of some variables on other variables

• Often for analysis, we use correlation and regression models

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 4


RESEARCH ISSUES (CONT’D)
• Often in research, we study the relationships between variables, or the
influences/ effects of some variables on other variables

• Often for analysis, we use correlation and regression models

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 5


RESEARCH ISSUES (CONT’D)
• Often we encounter research problem more sophisticated than just
relationships between TWO variables

• … Rather, we have several variables for independent, and for dependent too,
and…
• some dependent variables become independent variables to other variables

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 6


RESEARCH ISSUES (CONT’D)
• Thus, it might come across our minds, that our correlation which handles
two variables, or our regression model relating a single dependent to one (or
more) independent variables are no longer valid for a more sophisticated
research

• Let’s look at some of such sophisticated researches

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 7


EXAMPLE 1
• Single DV, Three Correlated IVs

REVISION

1
MATH SPM ACHIEVEMENT error

ATTEND LECT

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 8


EXAMPLE 2
• Multiple IVs, Multiple DVs, DV becomes IV for DVs

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 9


MORE COMMON, COMPLEX RESEARCH ISSUES
• The variables or factors that we are relating, are actually abstract concepts,
that have to be described or indicated by several (multiple) items or
indicators

• i.e. we are relating not directly measured variables

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 10


EXAMPLE 3
• Positive Symptoms influences Laterality. Both are unmeasured
concepts/factors, indicated by 3 and 2 items respectively.

• (

Kikuchi et al (2013)

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 11


EXAMPLE 4
• Relationships amongst unmeasured variables
E E E E

X1 X2 X3 X4

Price
E
H1
X5
E
X6 H2
Customer H4
Customer
Service
E
X7 Share Commitment
H3
E
X8
Atmosphere
Y1 Y2 Y3 Y4 Y5 Y6 Y7 Y8

X9 X10 X11 X12 E E E E E E E E

E E E E

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 12


EXAMPLE 5
1. QAProg affects
d1 .64 C1 .80
.58 zqt SEM Model
.26
d2 C2 .76 Chi-square = 350.743 (196)df
.65 .80 Cmindf =1.790 p-value = .000

Qteach which, in turn


d3 C3 QTeachr GFI = .936
.62 .79 .58 CFI = .976

1
d5 C5
TLI = .971

affects QTandL, and


RMSEA = .042
d18 .82
.68 C18
.87 .48
d19 .75 C19
in turn affects d20
.63 C20
.79
QAProg
.33 d11
.71
d12

QGrad. ALL are not 2


.27
zqtl .64
C11
.80 .66
C12 zqg .68
C33

C32
.47
d33

d32
.47

directly measured
.81 .56
.38 .75
.76 QTandL QGrad cgpa dcgp
.85 -.05 .71
.00
.82

variables/concepts.
.83 1.00 .04 GRADE dgrd
.41 .50
C10 C13 .00
.70 .67
d28 .79
.62 C28

2. QAProg correlates
.38 .26
.77 d10 d13
d30 .59 C30 .75 QSSuprt
d27 .56 C27

with QSSupport. -.35


d29 .80 C29 .89 d14 .82 C14
.91
QStud
d15 .69 C15
Both are indicated by .83 .55

dlrn LEARNST .15


.02 zqs

3 and 4 indicators
respectively
Nur Riza (2008)

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 13


RELATIONSHIP ANALYSES
• Various statistical techniques for analyzing various kinds or levels of
relationship analyses include:
o Correlation analysis
o Multiple regression
o Path analysis
o Factor analysis

• One analysis that can be used for the above objectives is an advanced
technique of
o Structural Equation Modeling

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 14


WHAT IS SEM
• Stands for STRUCTURAL EQUATION MODELING

a comprehensive statistical
approach to testing hypotheses a methodology for representing,
about relations among observed estimating, and testing a theoretical
and latent variables (Hoyle, 1995). network of (mostly) linear relations
between variables (Rigdon, 1998).

tests hypothesized patterns of directional and


nondirectional relationships among a set of
observed (measured) and unobserved (latent)
variables (MacCallum & Austin, 2000).

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 15


MAIN GOALS OF SEM
• 1) to understand the patterns of correlation among a set of variables and
• 2) to explain as much of their variance as possible with the model specified

• (Kline, 1998).

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 16


OTHER NAMES OF SEM
• Analysis of SEM is also known as
o Analysis of Covariance Structure
o Linear Causal Analysis
o Latent Variable Analysis
o Dependence Analysis
o Linear Structural Relations (LISREL

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 17


WHAT COMPRISE SEM?

Regression

Correlation Factor
analysis

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 18


RELATIONS BETWEEN VARIABLES
• Three types:
• 1) association e.g. correlation, covariance;
• 2) direct effect: directional relation between two variables, e.g. independent
and dependent variables; and
• 3) indirect effect: effect of an independent variable on a dependent variable
through one or more intervening or mediating variables (Hoyle, 1995).

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 19


FEATURES OF SEM
• Estimates two models simultaneously
o Measurement model
o Structural Model

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 20


MEASUREMENT MODEL
• specifies how latent variables are measured (using indicators i.e. observed
variables)
• Two measurement models – endogenous (DV) and exogenous variables (IV)
• i.e Factor Analysis.
l11 X1 d1

1
l21
X2 d2
WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 21
STRUCTURAL MODEL
• specifies how latent variables are related
• i.e. Regression or Path Analysis

1
2

1

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 22


DEPENDENT VS INDEPENDENT VARIABLES
• All variables are classified as either dependent or independent.
• Any variable that has a one-way path (or arrow) pointing at it is classified as a
dependent variable.
• All other variables are classified as independent variables.

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 23


ENDOGENOUS VS. EXOGENOUS
• ENDOGENOUS construct • EXOGENOUS construct
o Latent variable that are the o Latent variables that are the
‘dependent’ or ‘outcome’ in at ‘predictor’ or ‘causal’ for
least one causal relationship other construct. (Hair et al.,
(Hair et al., 2012). 2012)
o whose variation is to be o Whose variability is assumed
explained by exogenous and to be determined by causes
other endogenous variables in outside the causal model
the causal model under considerations
(Pedhazur,1997). (Pedhazur, 1997).
o Of ‘internal’ origin o Of ‘external’ origin
o Have one or more arrows leading o Have only causal arrows
INTO them leading OUT of them

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 24


OBSERVED VARIABLES
• Actual measures and scores.
• Manifest variables associated with
o exogenous constructs are labeled X.
o endogenous constructs are labeled Y.
• Might contain error, thus should include ‘measurement error’.
• indicators

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 25


ASSUMPTIONS OF SEM
• Observations are independent of each other
• Random sampling of respondents.
• Linearity of relationships between exogenous and endogenous variables
• Distribution of observed variables is multivariate normal.

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 26


TO SEM OR NOT SEM
• the research hypothesis being testing
• the requirement of sufficient sample size (e.g., 5-20 cases per parameter estimate)
• measurement instruments
• multivariate normality
• parameter identification
• outliers
• missing data
• interpretation of model fit indices
• (Schumacker & Lomax 1996).

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 27


APPROACHES TO SEM
• review the relevant theory and research literature to support model specification
• specify a model (e.g., diagram, equations)
• determine model identification (e.g., if unique values can be found for parameter
estimation; # df, for model testing is +ve)
• select measures for the variables represented in the model and collect data
• conduct preliminary descriptive statistical analysis (e.g. scaling, missing data,
collinearity issues, outlier detection)
• estimate parameters in the model
• assess model fit
• respecify the model if meaningful
• present and interpret the model.

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 28


Thank You!

Any Questions?

WORKSHOP SEM USING AMOS 27-28 JUNE 2016 NRMS 29

You might also like