EPSY 395c: Quantitative Research Design and Methods
Instructor:
Dr. Jianjun Wang
Education 149
Office Phone 654-3048
Office Hours: Friday
3:00-5:00 PM
Saturday
3:00-4:00 PM
E-Mail:
jwang@csub.edu
Text: (1)
Sample Surveys
http://www.eng.auburn.edu/users/teodoge/SAMPLE%20SURVEYS.doc
(2)
Repeated Measures Analysis (MANOVA vs. Mixed Modeling)
http://www.ats.ucla.edu/stat/sas/seminars/sas_repeatedmeasures/default.htm
(3)
New SAS Procedures for Analysis of Sample Survey Data
http://www.ats.ucla.edu/stat/SAS/library/svy_survey.pdf
http://www.ats.ucla.edu/stat/seminars/svy_intro/default.htm
(4)
Analyzing Longitudinal Data using Multilevel Modeling
http://www.ats.ucla.edu/stat/seminars/mlm_longitudinal/default.htm
(5)
Cluster Analysis and Discriminant Analysis (1 & 2)
http://psych-srv3.colorado.edu:16080/~carey/Courses/PSYC7291/
(6)
Canonical Discriminant Analysis (see reference book #1)
Reference Book/Resources
(1) Afifi, A. A. & Clark V. (1996). Computer-aided multivariate analysis (3rd edition).
New York: Chapman & Hall.
(2) Beside the web resources listed above, students are encouraged to conduct online
search of reference materials relevant to each of the topics covered in this class, and share
these additional documents during the classroom discussions.
Course Description:
An integrated approach has been taken in this class to cover (1) Different
techniques of data gathering under complex sampling framework; (2) Variable-orientated
data analyses pertaining to “narrow” (mixed modeling) and “wide” (MANOVA) formats
of the data array; (3) Subject-oriented data analysis for case classification and cluster
analysis.
The quantitative training has a three-fold focus: (1) when to use each statistical
method, (2) how to use the method through computer programming, and (3) how to
interpret the results in research literature and computer printout. At conclusion of the
course work, students will have a set of useful programs. By replacing numeric parts of
each program with their own data in the future, students will be able to conduct statistical
analyses, and produce empirical results for dissemination. The academic training is also
helpful to students who need to understand statistical results produced by other
investigators. Built on the multivariate statistical methods from EPSY 214: Intermediate
Statistics, this course will enrich student understanding of special considerations required
for quantitative inquiries that involve complex sampling and generalized linear modeling
approaches.
Course Grading
. There will be four quizzes covering the lectured contents and computing activities.
Those quizzes account for a total of 60% of the course grading (15% per quiz).
. Students are expected to understand characteristics of each statistical approach. More
specifically, students may choose three of the quantitative methods covered in this class,
and adduce examples of their applications in his/her particular work setting. Online
resources may be reviewed in support of the method selection and research design. The
description for each application should be concise, and confined within 3 pages. The
three essay discussions account for 30% of the course grading.
. Class participation (10%)
Tentative Course Schedule
Unit 1: Design and Analysis of Data from Complex Sampling
. Limitations of Simple Random Sampling
. Stratified Sampling
. Systematic Sampling
. Cluster Sampling
. Recent Advancement in Statistical Computing
Unit 2: Design and Analysis of Repeated Measures
. Needs and Designs of Repeated Measures
. Wide Form and Long Form of Data Array
. MANOVA vs. Mixed Modeling
. Analysis of Longitudinal Data using Multilevel Modeling
Unit 3: Design and Analysis of Messy Data
. Missing Treatment vs. Missing Observations
. Different Types of Sum of Squares
. Listwise vs. Pairwise Deletions
. Dummy Variable Adjustment
. Multiple Imputation
Unit 4: Subject-Oriented Research Design and Data Analysis
. Subject-Oriented vs. Variable-Oriented Studies
. Case Classification – Discriminant Analysis
. Discriminant Function Evaluation
. Cluster Analysis
. Evaluation of Distant Measures
Unit 5: Canonical Correlation Analysis
. Relationship with Multiple Regression and Correlation
. When to Use Canonical Correlation Analysis
. Canonical Variables and Canonical Loadings
. Redundancy Analysis
. Canonical Discriminant Function Analysis