Office: |
HH-110 |
Telephone: |
525-9140 x 23423 (24-hour Voice Mail) |
e-mail: |
Monday: 13:30, Wednesday: 10:30
Other times by appointment.
Don't hesitate to contact me by telephone, voice mail, or e-mail any time you need help. If you need to see me at any time and my office door is open, I will see you then if I can, or arrange a time to meet later.
The class will meet Monday 11:30-13:20 in BSB-103, Wednesday 11:30-12:30 in BSB-106. Some of that time will be spent in the JHE computer lab in a workshop format. Students will also be expected to devote time each week to independent study.
Statistical computing, statistical software packages, working with large data sets, exploratory data analysis, graphical methods, statistical consulting practice.
Learn statistical software packages, beginning with S-Plus, R and SAS, and continuing with any other systems the students wish to work with.
Learn exploratory data analysis, learn how to approach a large and complex data set.
Develop analytical skills using a broad range of statistical methods. Integrate and apply the techniques of statistical analysis learned in other courses. Improve understanding of linear models, generalized linear models and non-linear models and explore advanced multivariate methods.
Get some experience in consulting skills, and consultant-client interaction.
Learn how to document your work and write a report.
Statistical computing, statistical software packages, working with large data sets, exploratory data analysis, graphical methods, statistical consulting practice, linear and non-linear models, report writing.
We will work with the data sets chosen for the Case Studies Session at the Statistical Society of Canada 2004 Annual Meeting to be held at the University of Montreal, May 30 - June 2, 2004. We hope that this year, as in all past years, Statistics 4P03/6P03 students will be able to attend the meeting and present their results in the Case Studies Session. The Case Studies will be available from the Statistical Society of Canada web site by the end of January. Past Case Studies have been archived on that site.
- Chatfield, C. (1988) Problem Solving: A Statistician's Guide, Chapman & Hall.
This book is more suited to self-study than lecturing, but we will be discussing assigned readings. There are good examples showing how to begin analyzing different kinds of data sets. There is good advice on report preparation.
- Spector, P. (1994) An Introduction to S and S-Plus, Duxbury.
- Elliott, R.J. (2000) Learning SAS in the Computer Lab, Second Edition, Duxbury.
- Anonymous. (1992) UNIX: An Introductory User's Guide, CIS, McMaster University.
Also recommended:
- Cody, R.P. & Smith, J.K. (1997) Applied Statistics and the SAS Programming Language, Prentice Hall.
- Becker, R.A., Chambers, J.M. & Wilks, A. (1988) The New S Language, Wadsworth & Brooks/Cole.
- Chambers, J.M. & Hastie, T. (1992) Statistical Models in S, Wadsworth & Brooks/Cole.
You will need to refer to other manuals for Splus, SAS and any other software packages you may be using. There is a set of SAS manuals in the BSB Computer Lab. An Splus manual can be signed out from the Mathematics & Statistics office.
Some data sets may be chosen from:
- Johnson, R.A. & Wichern, D.W. (2002) Applied Multivariate Statistical Analysis, Fifth Edition, Prentice Hall.
- Rosner, B. A. (2000) Fundamentals of Biostatistics, Fifth Edition, Duxbury Press.
There will be three assignments. The assignments for 2003-2004 will be developed in class, in consultation with the students.
The four assignments listed below are from 1996-97. They are listed here only as examples. This year's assignments will follow a similar pattern.
- Choose one or both of the Case Studies prepared for the 1997 Annual Meeting of the Statistical Society of Canada, "Seal Vocalizations" and "Remote Sensing of Forest Crown Cover". Using Splus, apply EDA and submit a report showing what you have learned about the data. Your report should include a variety of graphs.
- Repeat your preliminary analysis of the SSC Case Studies, this time using SAS. Your first report will probably have indicated a number of directions worthy of further study and you should be able to improve on your earlier analysis in a number of ways. You may also be ready to try some model-fitting or multivariate analysis, if you consider it appropriate. Detail your results in a written report.
- Treat the crayfish size-frequency data introduced in class as a consulting experience. Write a plain-language report suitable for submission to the client.
- Use the HORMONE data set from Rosner. Apply linear and logistic regressions, as suggested in Problems 11.51-11.56, pp 545-546. Submit a written report.
All written reports must be submitted by the end of the regular undergraduate examination period.
Students will have an account on the Departmental UNIX machines. Splus, R and SAS will be available on redpine. They can be accessed from the BSB Windows computer lab, or from home, in teletype mode or with X-term emulators. Detailed instructions can be found in the Web document "Running the UNIX Versions of Splus, R and SAS at McMaster". Other software is available in the BSB Windows computer lab (SAS, MINITAB, SPSS, Excel, Splus, R). Students may install a licensed copy of Splus on their own computer. Students may download and install R on their own computers.
There will be no written tests or final examination.
The assignments will be worth a total of 80 marks. All will be judged on content, creativity, validity and accuracy. At least one assignment will also be judged on the quality of writing (25% of the total marks for the assignment).
Students may work in groups of two or three and submit assignments jointly, in which case each student will receive the same mark on the assignment.
The remaining 20 marks will be awarded for class participation; a student who attends all classes and contributes regularly to class discussion will receive full marks.
I will review all "borderline" marks and possibly make further adjustments.
Graduate students taking Statistics 6P03 for graduate credit must give a 30-minute class presentation and submit a written report on one of the Case Studies. The presentation should be of a standard appropriate for a scientific meeting.
This additional work will count for 10% of the graduate student's final mark. The mark computed from assignments, tests and examinations by the grading scheme shown above will count for 90%.
We remind you of the Statement on Academic Ethics and the Senate Resolutions on Academic Dishonesty found in the Senate Policy Statements distributed at registration and available in the Senate Office, also the School of Graduate Studies Statement Regarding Academic Ethics. Any student who infringes one of these resolutions will be treated according to published policy.