General Information :

• Syllabus (2011).
• Course logThis contains a record of what we cover in lecture, homework, etc., and a running plan for the next few week or so (based on last time I taught the course).

• April 6: Haiyan's office hour for Mon April 11 are canceled and are rescheduled for Wed April 13 3-5pm. If this make completing homework difficult, you may turn in your homework on Thursday, April 14.
• April 6: The take home final will be due Monday May 9, 5pm.
• March 8: The mid-term and data are posted (& links checked)
• Feb 2:
  • I hope that you are all staying home today. I fixed the link to homework #2. You should be able to work to it if you have internest access or SAS on your own computer. I'll be checking e-mail fairly regularly if you have any questions about it.
  
  
  • I will have put another set of up-dated notes (below) and will have time to do some more today.
• Jan 26: A bit of help on homework 1--- you need to compute the angle between two vectors, say x and y, assuming that you have computed the length of x and y (i.e., Lx and Ly), the SAS code (with comments) for this would be
  
  cos_x_y = x`*y/(Lx*Ly);
  
  * compute inverse of cos and answer given in terms of radians;
  rad_x_y = arcos(cos_x_y);
  
  * create a scalar equal to pi;
  pi = constant('Pi');
  
  * convert radians to degrees;
  degree_x_y = (180/pi)*rad_x_y;
  
  
  
• Jan 26: If you do not have the textbook, there is a scan of the page in the text that has the assigned problem for homework #1.
• An introduction to SAS will be given Wed Jan 26 at 9am in rm 176 Education. Bring a laptop. The material covered is after lecture materials on linear algebra. If you have used SAS there is no need to attend (i.e., this is optional).
• Jan 18:
  • Lecture meeting time: Lectures are from 10:00-11:50am..
  • Computing: We will be running statistical software (SAS) using remote desktop connection. We will connect to a server that has the software. There are 2 things that you need before you can do this:
• Your computer must be hooked up to the internet (campus, home, hotel, Espresso Royal, elsewhere).
• You must be registered for the class. If you reccently registered, you may not yet have permission to login.

• Introduction: (up-dated 2011)
  • Lecture notes.
  • SAS example: some graphics.
  • Using SAS assist for graphics.
  • Using SAS insight. (interactive data analysis)


• Linear Algebra: (up-dated 2011)
  • Lecture notes.
  • Chapter by Larry Hubert..
  • Link to online matrix reference mannual by Mike Brookes.
  • Link to Online Matrix Cookbook..
  • MATLAB presentation by Larry Hubert..
  • SAS information and example:
    • SAS PROC/IML example. IML is short for "interactive matrix language".
    • SAS IML documentation and what's new. This is also available online when you're running SAS.
  • R example
    • R script from lecture notes.
    • scores_data.txt. This is what the follow R code will look for.
    • basic_statistics_R.txt. Reading in data and computes sample statistics. This also shows how to write functions and use them.


• Optional: Introduction to SAS to be held in rm 176 Education bldg. We will go over the first two items.
  • When: Jan 26, 9am
  • Introduction to SAS notes..
  • hsb-data.sas.
  • SAS PROC/IML example. IML is short for "interactive matrix language". We went over part of this in lectre.
  • Using SAS assist for graphics.
  • Using SAS insight. (interactive data analysis)
  • Slick online SAS training from SAS.com
  
  
• Geometry of the Sample (up-dated 2011)
  • Lecture notes.


• Linear Combinations (up-dated 2011)
  • Lecture notes .


• Multivariate Normal Distribution (up-dated 2011)
  • Lecture Notes..
  • Bi-variate normal distribution demonstration. An interactive MATLAB program written by Peter K. Dunn that demonstrates the effect of changing mu, Sigma and rho.
  • R Script that graphs Bi-variate normal distribution. This is from http://www.stat.ucl.ac.be/ISpersonnel/lecoutre/stats/fichiers/_gallery.pdf, where you can find lots of other R scripts for graphics.
  • Sample output from R script.


• Inferences About a Mean Vector (up-dated 2011)
  • Lecture Notes.
  • SAS/IML: WAIS example.
  • SAS/IML: 4 Psychological tests example. (including how to read data into SAS/IML, and showing the relationship between T^2 and Wilk's lambda).


• Comparisons of Two Means (up-dated 2011)
  • Lecture Notes.
  • SAS: Exassy example in the lecture notes..
  • SAS: Airconditioning example.
  • SAS: Calculator example.(3/11/2008) Correction made to calculator sas code (i.e., module for computing Tsq for hypothses of form H_o: mu = mu_o where the mu's are vectors. There were TWO lines that needed to be corrected. The correct lines are
    dfden = n-p;
    F =(n-p)/((n-1)*p)*Ts;
    Note that the error is still in the lecture notes.
  • SAS/IML model for 2 independent T2 test.
  
  
  
• More Linear Algebra: eigensystems, SVD & maximization (up-dated 2011)
  • Lecture Notes.
  • Overview/Summary Handout.
  • SAS/IML: Two independent sample T test.
  • SAS: IML and Biplot.
  • Biplots in Practice by Michael Greenacre... for more information about bi-plots.
  • Least Squares Optimization in Multivariate Analysis by Jos M.F. ten Berge...for more information on linear algebra and applications to multiple regression, principal compenents analysis, simultaneous components analysis in two or more populations, MINRES factor analysis, canonical correlation, redunancy analysis, PARAFAC and INDSCAL.
  • For additional resources, see linear algebra notes above.
  
  
  
• Principal Components Analysis (up-dated 2011)
  • Lecture Notes.
  • Biplots in Practice by Michael Greenacre
  • Least Squares Optimization in Multivariate Analysis by Jos M.F. ten Berge...for more information on linear algebra and applications to multiple regression, principal compenents analysis, simultaneous components analysis in two or more populations, MINRES factor analysis, canonical correlation, redunancy analysis, PARAFAC and INDSCAL.
  • SAS: European Jobs.
  • SAS: Swiss Bank Notes
  • Picture of Swiss Bank Note (variables labeled aren't quite what we're using).
  • SAS: Men's Track Data
  • SAS: Scree Plots for all examples.
  
  
  
• One-Way MANOVA (up-dated 2011)
  • Lecture Notes.
  • SAS: Shorthand example.. This uses PROC GLM and in PROC/IML rank g design matrix.
  • SAS: Cameron-Pauling cancer data example. This uses PROC GLM, and in IML, rank (g+1) design matrix, traditional method. Computing various contrasts are illustrated here.
  
  
  
• Extensions of one-way MANOVA (up-dated 2011)
  • Lecture Notes.
  • SAS: WAIS and Eldery. (profile analysis example).
  • SAS: shorthand example revisited. (2-way MANOVA).
  • SAS: unbalanced design example.
  
  
  
• Discriminant Analysis (up-dated 2011)
  • Lecture Notes.
  • SAS:WAIS example. (2 group analysis).
  • SAS:Shorthand example: (6 groups) SAS PROCS GLM, IML, CANDISC, and DISCRIM.
  
  
  
• Canonical Correlation Analysis
  • Lecture Notes.
  • Link to Heck (1960). (Will only work if have uiuc account or jstor).
  • Link to Pillai & Bantegui (1959). (Will only work if have uiuc account or jstor).
  • SAS: WAIS example. (how to input a correlation matrix as a data set, PROC CANCOR).
  
  
  
• Factor Analysis
  • Lecture Notes. or an alternative format (but otherwise equivalent version)
  • SAS MACRO that does item analysis. This macro runs PROC CORR and uses PROC IML to test hypotheses on Cronbach's alpha and provides (1-alpha)x100% confidence intervals for alpha.
  • SAS Macro that does Item Analysis, Cronbach alpha CI & tests, Factor Analysis. (this macro uses PROCs CORR, IML and Factor).
  • SAS for example presented in class & more.


Homework & Exams:

• Homework writing guidelines..


• Homework #1
  • Homework assignment 1. (Due Wed Feb 2)
  • Page 38 from text (needed for one of the problems)
  • Data for Problem 1.4 from the text (including sas program that creates a sas data set).
  • SAS code for obtaining answers.
  • Answer key. (with typos corrected)
• Homework #2
  • Homework assignment 2. (due Wed Feb 9)
  • SAS/IML code that gives answers to all problems.
  • Answer key.
• Homework #3
  • Homework assignment 3 (due Wed Feb 16)
  • SAS/IML code that gives answers most of the problems.
  • Answer key.
• Homework #4
  • Homework assignment 4 (due Wed Feb 23)
  • SAS/IML code
  • Answer key
• Homework #5
  • Homework assignment 5 (due Wed March 2)
  • Broncial cancer data.
  • Answer key.
  • 4.26 from J&W.
  • 5.1 from J&W.
  • 5.2 from J&W.
  • Bronchical cancer problem.
• Homework #6
  • Homework assignment 6 (due March 9--- NO LATE ASSIGNMENTS)
  • hsb-data.sas (High School & Beyond Data).
  • About the High School & Beyond Data Set: A description of the variables and coding.
  • Answer Key.
  • SAS code.
• Mid-term Exam (will be available on or before March 9 and is due March 16th)
  • Mid-term exam.
    Please attach computer code as an appendix to your exam.
  • Data for problem 2.
  • Data for problem 3.
  • Data for problem 4.
• Homework #7 (due Wed April 6)
  • Homework assignment 7
  • Women's track data.
  • Answer Key.
  • Women's track data: SAS for solutions.
• Homework #8 (due Wed April 13)
  • Homework assignment 8.
  • Fish data.
  • Answer Key.
  • SAS: Fish data MANOVA.
  • SAS: Johnson & Wichern 6.8.
• Homwork #9 (due Wed April 20)
  • Homework assignment 9
  • Fish data.
  • Answer Key.
  • SAS: Profile analysis (problem 2).
  • SAS: Problem 3.
  • SAS: Problem 1.
• Homework #10 (due Wed April 27)
  • Homework assignment 10.
  • Fish data.
  • Answer Key.
  • SAS for discriminant analysis.


• Homework #11
  • Homework assignment 10. (Bring to class for factor analysis lecture)
  • Answers given in lecture.


• Final Exam (Due Monday May 9, 2011)
  • Final Exam
  • Data for problem 1.
  • Data for problem 2.
  • Data for problem 3.


Handy Programs and Links:
• CIforP.f: A FORTRAN program that computes large sample confidence intervals for a proportion. For PC computers, a execultable version of CIforP (i.e., already complied)
• pvalue.f: A FORTRAN program that computes p-values and (bonferroni) critical values for the standard normal, chi-squared, t, and F distributions (and for correlations). For users of PC type computers, pvalue.exe is an executable (i.e. already compiled) program.
• Link to R homepage.