BSAD 6314 Section 001
Multivariate Statistics
Fall 2010
Professor Contact Information
Wendy J. Casper, Ph. D.
Office: Business Building 233
E-mail: wjcasper@uta.edu
Phone: (817) 272-1133
Office Hours: By appointment
Course Time & Location:
BSAD 6314 Section 001; Tuesdays, 2:00 pm – 4:50 pm; COBA 251
Course Website:
http://management.uta.edu/Casper/MultiStat/bsad_6314.htm
Course Description:
This course is designed to help you to effectively apply, interpret, and evaluate different
multivariate statistical techniques. Throughout the course there will be an emphasis on both
conceptual understanding and the development of practical "how-to" skills. Topics covered
in the sequence are organized in terms of complexity, beginning with a broad overview,
moving into regression, and ending with structural equation modeling.
Business scholars have made use of a broad range of methods and analytical strategies to
address questions of interest. Because each approach to answering research questions
involves trade-offs, researchers have often found it necessary to employ a combination of
analytical techniques to reach any firm conclusions. A major goal of this course is to
facilitate decision making within these constraints.
We will discuss a variety of advanced statistical techniques. Throughout the semester, you
will gain hands-on experience through projects and learn how to draw statistical and
substantive conclusions from results of analyses. You will be asked to prepare written
summaries of results using either Academy of Management style or style guidelines for
other journals in your field. If you use style guidelines other than those of the Academy of
Management or the American Psychological Association, please provide a copy of these
guidelines to me along with your first project so I can refer to them as I evaluate your
projects.
Learning Objectives:
By the end of this course students will be able to:
1. Choose the appropriate multivariate technique to answer specific research questions
2. Describe the pros and cons of using various techniques to control for missing data
3. Describe the assumptions required to use various multivariate techniques including
regression, canonical correlation, logistic regression, discriminant analysis,
MAVOVA, factor analysis and structural equation modeling.
1
4. Use SPSS to run statistical analysis in regression, logistic regression, discriminant
analysis, and factor analysis.
5. Use LISREL to run a confirmatory factor analysis
6. Interpret the results of data analysis by examining SPSS output from regression,
canonical correlation, logistic regression, discriminant function analysis,
MANOVA, and factor analysis.
7. Interpret the resutls of LISREL analysis conducted to test the fit of both structural
and measurement models.
Required Text:
Hair, J. F., Black, W.C., Babin, B. J., Anderson, R.E., Tatham, R. L. 2006.
Multivariate Data Analysis (6th Edition) Upper Saddle River, NJ: Prentice Hall.
Recommended Supplemental Materials:
Academy of Management Journal Style Guide.
http://aom.pace.edu/amjnew/style_guide.html
Academy of Management Review Style Guide. http://aom.pace.edu/AMR/style.htm
SPSS Statistical Procedures Companion. Prentice Hall: Upper Saddle River, NJ.
Version 14 or later.
Course Requirements:
Course requirements include: (1) four project assignments and brief write-ups of results in
appropriate format; (2) Identification of articles using techniques discussed in class; (3) a
midterm exam; (4) a final exam; and (5) a final paper.
Grades are determined as follows:
40% Evaluation of the assigned projects (10% each)
5% Article Identification
15% Midterm exam
15% Final exam
25% Final paper
Final grades: A (90%); B (80%); C (70%); D (60%); F (below 60%)
Article Identification: Specific disciplines often emphasize different techniques, and your
analytic choices will often be evaluated based on norms of use in your discipline. Thus,
after the completion of each multivariate technique in class, you will be required to identify
examples of how these techniques are used and written up in top journals in your discipline.
Each week that article examples are due, you should (1) bring enough copies of articles to
provide a copy of the article to myself and all class members and (2) be prepared to get up
2
in front of the class and give a brief presentation of the content of your article including (a)
variables examined, (b) analyses conducted, and (c) results of analyses.
Final Paper: The final paper should examine research questions or hypotheses germaine
to your discipline from a set of data. You must use one or more of the multivariate
techniques that we have covered in class to conduct your analysis. Results must be
written in AMJ style or style guidelines in your discipline. Management students will be
required to submit their final paper to the 2005 Academy of Management conference.
Students in academic programs other than Management will be required to submit their
papers to a national conference in their academic discipline. In order to ensure you are
on track with the paper early in the semester, you will be required to hand in an outline of
what you plan to do one month into the semester. The outline should address the
following:
1.
What data set will you use? How did you obtain access to this data? (Please
ensure all approvals to use data have been obtained prior to handing in your
outline) What is your sample size? What variables are included in this data
set?
2.
What are the hypotheses or research questions you plan to answer? (Have at
least 3 hypotheses or questions).
3.
What multivariate statisical techniques will you use to test each
hypothesis/answer each question? (You will need to ensure you meet
assumptions associated with each technique).
If you do not already have datasets to use from your own research projects or
collaborations with another professor, I have several datasets which could be used for
this project. If you are interested in this, please see me to discuss. If you begin
working on a data set that I provide you, I would expect that you and I would
continue to collaborate on the paper for publication if there are interesting findings.
If you use a data set provided to you by another professor or colleague, it is a good
idea to discuss expectations regarding ongoing research up front with whoever
provides the data to you.
CARMA Webcasts: You will be required to watch several different webcasts from the
Center for Advancement of Research Methods and Analysis (CARMA) as part of your
course requirements. The webcasts will sometimes be offered at times other than the
scheduled class time. If you can not attend when the CARMA webcasts are offered, you
may access these webcasts on your own time via the web and watch on your own.
However, you are responsible for the material in the webcasts and there may be questions
from the webcasts on the exams. I am only asking you to attend the webcasts that deal
with statitical issues regatding multivariate techniques, so all material in the webcasts
should be relevant to the course.
3
Class Policies
1. Attendance. As graduate students, I expect that you all will attend class and be
engaged in learning. However, if you miss class, you are responsible for class
material and announcements made in class including changes to the syllabus.
2. Class Disruptions. Please come to class on time, turn off your cell phones and pagers
before class, and refrain from other activities that disrupt class.
3. Due Dates. Projects, article examples and the final paper must be handed in by the
beginning of class on the day they are due. Anything that is handed in late will
receive 50% of the possible points if handed in within a day. No points will be given
for any work that is handed in more than one day late.
4. Collaborative Work. You are encouraged to work together to conduct data analysis
for projects and prepare for exams. However, each student must independently write
up the results of the data analysis for projects and hand it in with printouts from
analyses. Students should complete exams and the final paper without assistance from
other students.
5. Make Ups. Please do not miss exams. Make-up exams will not be allowed except
under conditions of documented severve illness or emergency.
8. Academic Integrity. Students are responsible for maintaining academic integrity.
Cheating on exams or plagiarism of assignments or papers from other students or
published sources (including the internet) is a violation of academic integrity and
professional ethics. Cheating includes handing in work (either your own or others’)
for this class that was completed for another class. Dishonesty in reporting results or
unethical behavior in research is also a violation of academic integrity. Engaging in
any behavior that violates academic integrity can result in failure of the course and/or
other penalties.
4
CLASS SCHEDULE
This course is a dynamic process, subject to change. You are responsible for maintaining
awareness of changes in class scheduling if you have missed class.
Date
Topic
Aug 31
Overview of Multivariate Statistics
Read:
Hair, Chapter 1
Bobko, P. 1990. Multivariate Correlational Analysis. In Dunnette,
M. D. & Hough, L. M. (Eds.) Handbook of Industrial/Organizational
Psychology (2nd Edition), Volume 1, pp. 637-686. Palo Alto, CA:
Consulting Psychologists Press.
Sept 7
Data Cleaning and Multivariate Techniques
Read:
Hair, Chapter 2
Orr, J. M., Sackett, P. R., & Dubois, C. L. Z. 1991. Outlier detection
and treatment in I/O psychology: A survey of researcher beliefs and an
empirical illustration. Personnel Psychology, 44: 473-486.
Roth, P.L., & Switzer, F.S. 1995. A monte carlo analysis of
missing data techniques in a HRM setting. Journal of Management, 21:
1003-1023.
Sept 14
Multiple Regression
Read:
Hair, Chapter 4
St. John, C. H. & Roth, P. L. 1999. The impact of cross-validation
adjustments on estimates of effect size in business policy and strategy
research. Organizational Research Methods, 2: 157-174.
5
Sept 21
Regression: Mediation & Moderation
DUE article using multiple regression
Project 1 Assigned
Read:
Aiken, L. S., & West, S. G. 1991. Multiple regression: Testing and
interpreting interactions (Ch. 2-4, pp. 9-61; Ch. 7, pp. 116-138). Newbury
Park, CA: Sage.
Baron, R. M., & Kenny, D. A. 1986. The moderator-mediator
variable distinction in social psychological research: Conceptual, strategic,
and statistical considerations. Journal of Personality and Social
Psychology, 51: 1173-1182.
Sept 28
Canonical Correlation
DUE article using mediated regression
DUE article using moderated regression
DUE Project 1
Read:
Hair, Chapter 8 from 5th edition (see on website of readings)
Extra Sept
Logistic Regression: Dichotomous Dependent Variables
DUE article using canonical correlation
Read:
Hair, pp. 269-275; 366-382
Green, G. H., Boze, B. V., Choundhury, A. H., & Power, S. 1998.
Using logistic regression in classification. Marketing Research: 5-31.
Huselid, M. A., & Day, N. E. 1991. Organizational commitment, job
involvement, and turnover: A substantive and methodological analysis.
Journal of Applied Psychology, 76: 380-391.
6
Extra Sept
Discriminant Analysis
DUE article using logistic regression
DUE article using disrciminant analysis
Project 2 Assigned
Read:
Hair, pp. 276-355
Betz, N. E. 1987. Use of discriminant analysis in counseling
psychology research. Journal of Counseling Psychology, 34: 393-403.
Sept 30
DUE Project 2
Oct 5
Midterm Exam In Class
Oct 19
MANOVA
DUE Outline of Final Paper. Include data set choosen, hypotheses to
test and technique to use to test them
Read:
Hair, Chapter 6
Haase, R. F., & Ellis, M. V. 1987. Multivariate analysis of
variance. Journal of Counseling Psychology, 34: 393-403.
Oct 26
Factor Analysis I
Read:
Hair, Chapter 3
Conway J. M., & Huffcutt A.I. 2003. A review and evaluation of
exploratory factor analysis practices in organizational research.
Organizational Research Methods, 6: 147-168.
Ford, J. K., MacCallum, R. C., & Tait, M. 1986. The application of
exploratory factor analysis in applied psychology: A critical review and
analysis. Personnel Psychology, 39: 291-314.
7
Nov 2
Factor Analysis II
DUE article using exploratory factor analysis
Project 3 Assigned
Read:
Hurley, A. E., Scandura, T. A., Schriesheim, C. A., Brannick, M. T.,
Seers, A., Vandenberg, R. J., & Williams, L. J. 1997. Exploratory and
confirmatory factor analysis: Guidelines, issues, and alterations. Journal of
Organizational Behavior, 18: 667-683.
Nov 9
Structural Equation Modeling Part I
DUE Project 3
Read:
Hair, Chapter 10 and 11
Anderson, J. C., & Gerbing, D. W. 1988. Structural equation
modeling in practice : A review and recommended two-step approach.
Psychological Bulletin, 103: 411-423.
Mulaik, S. A., James, L. R., Van Alstine, J., Bennett, N., Lind, S.,
& Stilwell, C. D. 1989. Evaluation of goodness-of-fit indices for
structural equation models. Psychological Bulletin, 105: 430-445.
Nov 16
Structural Equation Modeling Part II
DUE article using structural equation modeling
Project 4 Assigned
Read:
Hair, Chapter 12
Nov 23
DUE Project 4
Dec 7
Final Exam in Class
DUE Final Paper
8
SUPPLEMENTAL READINGS BY TOPIC AREA
Data Cleaning
Roth, P. L., Switzer, F. S., III, & Switzer, D. 1999. Missing data in multiple item
scales: A Monte Carlo analysis of missing data techniques. Organizational Research
Methods, 2: 211-232.
Smith, P. C., Budzeika, K. A., Edwards, N. A., Johnson, S.M., & Bearse, L. N.
1986. Guidelines for clean data: Detection of common mistakes. Journal of Applied
Psychology, 71: 457-460.
Allison, P. D. 2002. Missing data. Thousand Oaks, CA: Sage.
Multiple Regression
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. 2003. Applied multiple
regression/correlation analysis for the behavioral sciences (3rd Edition). Mahwah, NJ:
Lawrence Erlbaum.
Darlington, R. B. 1968. Multiple regression in psychological research and
practice. Psychological Bulletin, 69: 161-182.
Mediated and Moderated Regression
Aguinis, H., & Stone, E. F. 1997. Methodological artifacts in moderated multiple
regression and their effects on statistical power. Journal of Applied Psychology, 82: 192206.
Boal, K. B., & Bryson, J. M. 1987. Representation, testing and policy
implications of planning processes. Strategic Management Journal, 8: 211-231.
MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V.
2002. A comparison of methods to test mediation and other intervening variable effects.
Psychological Methods, 7: 83-104.
Useful aid for graphing interactions:
http://www.jeremydawson.co.uk/slopes.htm
Canonical Correlation
Harris, R. J. 1989. A canonical cautionary. Multivariate Behavioral Research, 24:1739.
Logistic Regression
Menard, S. 2002. Applied Logistic Regression Analysis. Thousand Oaks, CA:
Sage.
9
Discriminant Function Analysis
Hsu, L. M. 1989. Discriminant analysis. A comment. Journal of Counseling
Psychology, 36:244-247.
O’Gorman, T. W. & Woolson, R. F. 1991. Variable selection to discriminate
between two groups: Stepwise logistic regression or stepwise discriminant analysis? The
American Statistician, 45: 187-193.
MANOVA
Huberty, C. J. & Morris, J. D. 1989. Multivariate analysis versus multiple
univariate analyses. Psychological Bulletin, 105, 302-308.
O’Brien, R. G., & Kaiser, M. K. 1985. MANOVA method for analyzing repeated
measures designs: An extensive primer. Psychological Bulletin, 97, 316-333.
Structural Equation Modeling
Bagozzi, P. P. & Yi, Y. 1988. On the evaluation of structural equation models.
Academy of Marketing Science, 16: 74-94.
Cortina, J. M., Chen, G., & Dunlap, W. P. 2001. Testing interaction effects in
LISREL: Examination and illustration of available procedures. Organizational Research
Methods, 4: 324-360.
Hall, R. J., Snell, A. F., & Foust, M. S. 1999. Item parceling strategies in SEM:
Investigating the subtle effects of unmodeled secondary constructs. Organizational
Research Methods, 2: 233-256.
MacCallum, R. C., Roznowski, M., & Necowitz, L. B. 1992. Model modifications
in covariance structure analysis: The problem of capitalization on chance. Psychological
Bulletin, 111: 490-504.
McDonald, R. P., & Ho, M. R. 2002. Principles and practice in reporting
structural equation analyses. Psychological Methods, 7: 64-82.
10