to - Jonathan Templin`s Website

Transcription

to - Jonathan Templin`s Website
ERSH 8750, Spring 2012
Introduction to Structural Equation Modeling Syllabus
Syllabus
Professor
Jonathan Templin
570B Aderhold Hall
[email protected]
706-680-7148
Course Information
Wednesdays: 1:25 pm – 4:25 pm
Room 119 Aderhold Hall
Prerequisite
ERSH 8310 (ANOVA) and 8320 (Regression)
and ERSH 8610 (Educational Measurement)
Office Hours
Tuesdays: 11:00am-1:00pm or by appointment
Note: Office hours are held in 228 Aderhold
(computer lab on 2nd floor)
Course Website
http://goo.gl/kaWyI
Course Discussion Page
http://goo.gl/lOMNb
Introduction to structural equation modeling is a course that seeks to teach students about the
fundamentals of structural equation modeling, combining theoretical and practical perspectives.
The course is designed to provide details of structural equation modeling, from the statistical
underpinnings to how to run many various types of structural equation analyses.
Course Objectives
Overall, this course is a course that teaches multivariate statistical thinking, structural equation
modeling, and the language of both methods. By the end of the course, students are expected to:



Understand the types of hypotheses and research questions for which structural equation
modeling is used
Know and perform structural equation modeling techniques using Mplus
Understand how structural equation models fit into a larger framework of statistical methods
Be advised: this course will challenge you, and will require a significant commitment, both in
amount of time and in amount of work. Expect to spend 9 - 12 hours outside of class each
week on this course. Reading the assigned papers and chapters in advance of lecture,
completing the homework each week, and attending class are keys to your success.
Required Textbook
None- Readings will be assigned and administered each week.
Prerequisite
This course assumes you have taken at the very least one year of Ph.D. level statistics
coursework. Within the College of Education at the University of Georgia, this means having
taken and completed both ERSH 8310 (ANOVA), ERSH 8320 (Regression), and ERSH 8610
(Educational Measurement). If you have not taken these courses (or similar courses), you will be
at a significant disadvantage which will likely slow you learning of the material and may slow
the class as a whole, including students who have taken the prerequisites. The first homework
(assigned the first day of class and due next Wednesday) is designed to test your knowledge of
the prerequisites.
If you:




Have not taken the prerequisite courses
Find the first day difficult
Find the homework difficult
Are not able to commit the time it will take to successfully complete this course
If any of these points apply to you, I strongly urge you to consider dropping this course as
there are a number of people hoping to add the course who have taken the prerequisites.
Statistical Computing
This course will Mplus for all analyses. Mplus, a powerful generalized modeling package, is
available to you in two ways:
1. You can purchase your own copy of Mplus through the Muthen & Muthen website:
https://www.statmodel.com/orderonline/categories.php?category=MplusSoftware/Student-Pricing/Click-here-to-order-download-only
2. The computer lab in room 228 has Mplus installed on all windows computers
Aderhold computer labs are open on a varying schedule throughout the semester. Please note:
Aderhold labs are not open during weekends and the schedule is subject to change. Purchasing
Mplus may be the best option to ensure you can complete all coursework.
Course Website/Technology
This course will not use the e-Learning Commons technology. Instead, we will use freely
available commercial software for communication and dissemination of course materials. Please
do not use eLC to email me as I will not get your message (use [email protected] instead).
Audio Recordings of Class
I will be making an MP3 recording of each class, which will be posted on the website by the end
of the day following class.
Class Page on Facebook
In order to facilitate communication, I have created a Facebook page where I will:
1. Post information about the course
2. Hold online office hours (as needed)
3. Answer your questions directly using the wall feature
The Facebook page is designed to facilitate discussion in a manner that is easily accessible by all
students and auditors of the course. To be a part of the discussion, please search for ERSH 8750
or use the link: http://goo.gl/lOMNb. To follow, you must “like” the page.
Course Materials Over Dropbox
All course readings will be available over a shared folder on Dropbox, a file repository online
(www.dropbox.com). To gain access to the shared folder, please send me an email. You do
not have to install the Dropbox application as you can download files from any web browser.
If you do not have a Dropbox account, please email me for an invitation.
Course Website
Course lecture slides (if available), course audio files, course assignments, and course
information is available on the website.
The website for the course is http://goo.gl/kaWyI
Course Structure and Student Evaluation
Student evaluation will be made on the basis of homework grades. All homework and answers
must be your own and not be copied or paraphrased from anyone else’s answers. You are
responsible for your own work.
Homework
Weekly homework assignments will be administered in order to give students practice applying
techniques discussed in class and will be due at the start of class the following week. Each
assignment must be at least 75% complete in order to be accepted for grading. Homework must
be submitted electronically (email [email protected]) in the form of Microsoft Word
document with the name: 8750_FirstLast_HW#.docx. Late homework will have a penalty of
10% of the total per calendar day (any homework later than 10 days late will not be accepted).
Do not wait until the last minute to do your homework.
Note: due to the size of the course (expected to be 30-40 people), grading will mostly be on a
correct/incorrect basis. Homework solutions will be posted each week after the last homework
assignment has been submitted.
Course Grading System
Percentage of Points
100-93
92-90
89-87
86-83
82-80
79-77
76-73
72-70
69-60
Below 60
Grade
A
AB+
B
BC+
C
CD
F
Academic Honesty
All students are expected to abide by the University of Georgia student honor code. You can
view the UGA academic honesty policy at http://www.uga.edu/honesty/.
Course Style and Content
Lecture Format
Most lectures will have notes (slides) available digitally, with slides available online by the
morning of the day of the lecture. Please check the course website before coming to class if you
would like to bring a printout of the slides with you. If nothing is posted, then we will be having
lecture without slides. I strongly encourage you to participate in lecture by asking questions
whenever anything is unclear.
Reading Assignments
To be fully successful in this course, I strongly encourage you to read the assigned papers and/or
chapter(s) prior to the coming to class when we will cover the topic. Even if you have difficulty
reading the material, exposure to the information prior to lecture will aid in your understanding
of the course. Remember, this course is about learning the language of structural equation
modeling and multivariate statistics.
How to Succeed in this Course
 Read the assigned papers (even if it doesn’t make sense to you – it will eventually)
 Come to class (and bring your questions about what you just read that week)
 Ask questions when you do not understand
 Come to office hours – we will meet in the lab where you can ask about MPlus
 Do the homework (consider it practice on applying statistics)
 Compare your homework with the solutions online before receiving your feedback
Tentative Course Schedule (subject to change as necessary)
Based on the topics of the course, we will have roughly five sections:
1.
2.
3.
4.
5.
Introduction and background information
Regression extensions/Path analysis/Simultaneous Equations
Confirmatory Factor Analysis
Structural Equation Models
Advanced Topics
Month
January
Day Week
11
1
January
18
2
January
February
February
February
February
February
March
25
1
8
15
22
29
7
3
4
5
6
7
8
9
March
March
14
21
10
March
April
28
4
11
12
April
April
11
18
13
April
25
14
Topic(s)
Introduction; Review of Regression and ANOVA;
Introduction to Mplus
Introduction to Matrix Algebra/Multivariate Normal
Distribution
Introduction to Maximum Likelihood and Missing Data
Path Analysis
Confirmatory Factor Analysis #1: Concepts/Identification
Confirmatory Factor Analysis #2: Fit Assessment
Confirmatory Factor Analysis #3: Theoretical Definitions
Multiple Group Factor Analysis
Introduction to Structural Equation Modeling:
Structural Models
NO CLASS: SPRING BREAK
Introduction to Structural Equation Modeling: Path
Analysis with Latent Variables
Advanced Topics: Mediational Analyses
Advanced Topics: Mean Structures/“Latent Curve
Modeling”/ SEM as Mixed Models
NO CLASS: AERA/NCME CONFERENCE
Advanced Topics: Estimation and Robust Estimation in
CFA/SEM
Advanced Topics: SEM with Non-Normal Data
Section
1
1
1
2
3
3
3
3/4
4
4
5
5
5
5
Course Reading List
Month
January
Day
11
Week
1
Readings
Mplus introduction website:
http://www.ats.ucla.edu/stat/mplus/seminars/IntroMplus/default.htm
Chapter 1: Introduction. Kline (2005). Principles and practice of
structural equation modeling (2nd Ed.). New York: Guilford.
Chapter 1: Historical foundations of structural equation modeling for
continuous and categorical latent variables. Kaplan, D. (2009).
Structural equation modeling: foundations and extensions (2nd Ed.).
Thousand Oaks, CA: Sage.
January
18
2
Chapter 2: Matrix algebra and random vectors. Johnson, R. A. &
Wichern, D. W. (2002). Applied Multivariate Statistical Analysis (5th
Ed.). Upper Saddle River, N.J., Prentice-Hall. (p. 84-100).
Chapter 4: The multivariate normal distribution. Johnson, R. A. &
Wichern, D. W. (2002). Applied Multivariate Statistical Analysis (5th
Ed.). Upper Saddle River, N.J., Prentice-Hall. (p. 149-177).
January
25
3
Chapter 1: An introduction to missing data. Enders, C. K. (2010).
Applied Missing Data Analysis. New York: Guilford.
Chapter 2: Traditional methods for dealing with missing data.
Enders, C. K. (2010). Applied Missing Data Analysis. New York:
Guilford.
Chapter 3: An introduction to maximum likelihood estimation.
Enders, C. K. (2010). Applied Missing Data Analysis. New York:
Guilford.
Chapter 4: Maximum likelihood missing data handling. Enders, C.
K. (2010). Applied Missing Data Analysis. New York: Guilford.
February
1
4
Chapter 5: Introduction to path analysis. Kline (2005). Principles
and practice of structural equation modeling (2nd Ed.). New York:
Guilford.
Chapter 6: Details of path analysis. Kline (2005). Principles and
practice of structural equation modeling (2nd Ed.). New York:
Guilford.
February
8
5
Chapter 7: Measurement models and confirmatory factor analysis.
Kline (2005). Principles and practice of structural equation
modeling (2nd Ed.). New York: Guilford
Chapter 3: Introduction to confirmatory factor analysis. Brown, T. A.
(2006). Confirmatory factor analysis for applied research. New
York: Guilford.
Chapter 4: Confirmatory factor analysis. Raykov, T., & Marcoulides,
G. A. (2006). A first course in structural equation modeling (2nd
Ed.). New York: Taylor & Francis.
February
15
6
Chapter 4: Specification and interpretation of confirmatory factor
models. Brown, T. A. (2006). Confirmatory factor analysis for
applied research. New York: Guilford.
Chapter 5: Confirmatory factor analysis model revision and
comparison. Brown, T. A. (2006). Confirmatory factor analysis for
applied research. New York: Guilford.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in
covariance structure analysis: Conventional criteria versus new
alternatives. Structural Equation Modeling, 6, 1-55.
February
22
7
Chapter 3: Factor analysis. Kaplan, D. (2009). Structural equation
modeling: foundations and extensions (2nd Ed.). Thousand Oaks,
CA: Sage.
Chapter 6: Reliability. Raykov, R. & Marcoulides, G. A. (2011).
Introduction to psychometric theory. New York: Routledge.
Chapter 7: Procedures for estimating reliability. Raykov, R. &
Marcoulides, G. A. (2011). Introduction to psychometric theory.
New York: Routledge.
February
29
8
Chapter 11: Multi-Sample SEM. Kline (2005). Principles and
practice of structural equation modeling (2nd Ed.). New York:
Guilford.
Chapter 7: Confirmatory factor analysis with equality constraints,
multiple groups, and mean structures. Brown, T. A. (2006).
Confirmatory factor analysis for applied research. New York:
Guilford.
March
7
9
Chapter 4: Structural equation modeling in single and multiple
groups. Kaplan, D. (2009). Structural equation modeling:
foundations and extensions (2nd Ed.). Thousand Oaks, CA: Sage.
Chapter 8: Other types of confirmatory factor analysis models:
higher-order factor analysis, scale reliability evaluation, and
formative indicators. Brown, T. A. (2006). Confirmatory factor
analysis for applied research. New York: Guilford.
Chapter 8: Models with structural and measurement components.
Kline (2005). Principles and practice of structural equation
modeling (2nd Ed.). New York: Guilford.
Boomsma, A. (2000). Reporting analyses of covariance structures.
Structural Equation Modeling, 7(3), 461-483.
March
14
March
21
NO CLASS: SPRING BREAK
10
Chapter 5: Structural regression models. Raykov, T., & Marcoulides,
G. A. (2006). A first course in structural equation modeling (2nd
Ed.). New York: Taylor & Francis.
DeShon, R. P. (1998). A cautionary note on measurement error
corrections in structural equation models. Psychological Methods, 3,
412-423.
McDonald, R. P., & Ho, M.-H. R. (2002). Principles and practice in
reporting structural equation analyses. Psychological Methods, 7, 6482.
March
28
11
MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). Mediation
analysis. Annual Review of Psychology, 58, 593-614.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G.,
& Sheets, V. (2002). A comparison of methods to test the
significance of the mediated effect. Psychological Methods, 7, 83104.
Edwards, J. R., & Lambert L. S. (2007). Methods for integrating
moderation and mediation: A general analytical framework using
moderated path analysis. Psychological Methods, 12, 1-22.
James, L. R.,Mulaik, S. A., & Brett, J. M. (2006). A tale of two
methods. Organizational Research Methods, 9, 233-244.
April
4
12
Chapter 6: Latent change analysis. Raykov, T., & Marcoulides, G. A.
(2006). A first course in structural equation modeling (2nd Ed.). New
York: Taylor & Francis.
Chapter 10: Mean structure and latent growth models. Kline (2005).
Principles and practice of structural equation modeling (2nd Ed.).
New York: Guilford.
Hancock, G. R. (1997). Structural equation modeling methods of
hypothesis testing of latent variable means. Measurement and
Evaluation in Counseling and Development, 30, 91 - 105.
Hancock, G. R., Kuo, W-L., Lawrence, F. R. (2001). An illustration
of second-order latent growth models. Structural Equation
Modeling, 8, 470-489
April
11
April
18
NO CLASS: AERA/NCME CONFERENCE
13
Chapter 1: Introduction (p. 22-37). Raykov, T., & Marcoulides, G.
A. (2006). A first course in structural equation modeling (2nd Ed.).
New York: Taylor & Francis.
Chapter 5: Improving the accuracy of maximum likelihood analyses.
Enders, C. K. (2010). Applied Missing Data Analysis. New York:
Guilford.
April
25
14
Chapter 1: The omni-presence of latent variables. Skrondal, A. &
Rabe-Hesketh, S. (2004). Generalized latent variable modeling:
multilevel, longitudinal, and structural equation models. Boca
Raton, FL: Chapman & Hall.
Chapter 2: Modeling different response processes. Skrondal, A. &
Rabe-Hesketh, S. (2004). Generalized latent variable modeling:
multilevel, longitudinal, and structural equation models. Boca
Raton, FL: Chapman & Hall.
Chapter 9: Data issues in confirmatory factor analysis: missing, nonnormal, and categorical data. Brown, T. A. (2006). Confirmatory
factor analysis for applied research. New York: Guilford.