Syllabus
CSC 423/324 - Data Analysis and Regression

Joseph Morgan
Office: CS&T 210
Phone: (312)-362-6321
Fax: (312)-362-6116
Email: jmorgan@cs.depaul.edu

Texts:

• An Introduction to Statistical Methods and Data Analysis. Ott. Duxbury Press, 4th edition, 1993.
• SAS System for Elementary Statistical Analysis. Schlotzhauer & Littell. SAS Institute, 2nd edition, 1991.
• Computer Science Data Analysis Using SAS. Knafl (unpublished notes). DePaul bookstore, June, 1997.

Topics:

1. Inferences about a population mean, about a difference in population means, using paired or independent samples, and about normality.
2. Introduction to analysis of variance. Multiple comparisons.
3. Multiple regression. Residual plots and regression diagnostics. Model selection using PRESS statistic.
4. Matrix notation and the general linear model.

The primary emphasis of the course will be on how to perform selected data analyses using the Statistical Analysis System (SAS). SAS PROCs to be covered include ANOVA, CORR, FREQ, GLM, IML, MEANS, NPAR1WAY, PLOT, PRINT, RANK, REG, SORT, STEPWISE, TTEST, and UNIVARIATE. Statistical and mathematical issues will be covered that assist the student in understanding the models underlying the statistical PROCs and in interpreting the output of those PROCs.

Grading:

Students grades will be based on the total points earned. Points will be assigned to homework problems, programming assignments, and examinations.

The grade breakdown will be as follows:

• Homework Problems: 15%
• Programming Assignments: 40%
• Final: 45%

Assignments:

Homework problems will either be assigned from the text or will be made up problems.

Programming assignments will involve analysis of a supplied dataset and will require a written analysis of the results using a word processing system. Programs should use meaningful names, titles, comments, and indentation. Analyses must be written in precisely stated statistical terms and in correct English. You should also use a word processor that supports mathematical notation and Greek symbols and you should integrate appropriate output from your SAS programs into your documents in a presentable format.

Assignments will be assessed penalties for late submission. Solutions to homework problems may be submitted by email or may be turned in to me in class or in my office, put into my mailbox, or faxed to me. All assignments must be submitted by the due date to avoid late penalties.

Plagiarism/Cheating:

Incompletes:

Readings:

1. The first set of topics to be covered in CSC 423/324 is related to the following sections of the texts.
   • Ott:
     • Chapters 1-7. Skip Sections 3.5, 4.4-4.8, 4.13, 5.3-5.4, 5.6, 6.4, 6.7 and 7.3. You should have covered much of this material, but perhaps it was presented differently. Read this material to gain a clearer understanding of the topics we are covering. The primary issues relevant to the class are those related to inference or hypothesis testing, and sampling distributions.
     • Relevant topics include descriptive statistics, one sample and two sample inference problems (both parametric and nonparametric approaches).
   • Schlotzhauer & Littel:
     • Chapters 1-7, Appendices 1-2, 4.
2. The second (analysis of variance) and third (regression and correlation) topics to be covered are related to the following sections of the texts.
   • Ott:
     • Chapters 9-15. Skip Sections 10.7, 12.6, 15.4, and 15.6. Multiple regression and analysis of variance topics will be covered. As much of the material in these Chapters will be covered as time allows.
   • Schlotzhauer & Littel:
     • Chapters 8-10.