Information Pertaining to Review for Final Exam

Note: You should plan to take the final exam at the time it is scheduled for --- it will not be easy for you to convince me to give you the exam at any other time.

The final exam is an open book(s) / open notes exam --- you can use whatever books and other printed or written material that you want to bring with you. (I don't think you should need anything other than the two text books and whatever notes that you've taken in class. I suppose that some may find printouts of some of my web page notes useful, but you won't want to spend a lot of time searching through those --- better to try to study the relevant parts prior to the time of the exam.) There is no need to use any tables pertaining to tests or distributions, SPSS, or even a calculator during the exam.

Extra Office Hours

I plan to hold extra office hours in our regular classroom (so that many can attend at the same time) during the following times: If you think that you will have a lot of questions, let me know ahead of time so I can start the STAT 535 office hours earlier. (I have sessions scheduled for my other class after your scheduled sessions on both days.) In the past, few 535 students have made use of my extra office hours, and so I didn't schedule a lot of extra time this year.

Study Guide

Although I may not have time to actually make the final exam until the two or three days right before the exam, I have now done enough planning so that I can let you know a lot of information about the nature of the final exam.

The exam will have 5 problems on it, with these problems having a total of 15 parts, with the point values for the parts ranging from 1.5 points to 4 points. (The total point value of the final exam will be 30, since in the end the final contributes 30% towards your overall score for the class.)

I'm giving you a lot of hints about what will be on the exam below. Some of the hints are intentionally a bit vague. Consider the vague hints to be a gift, and please do not ask for additional hints about what will or won't be covered, or for clarification concerning the hints given below. I will respond to very specific questions that you may have, but don't expect me to be too chatty if you say things like "Tell me what I should know about blah, blah, blah" or "Please review the important points of yadda, yadda, yadda," where yadda, yadda, yadda corresponds to one of the topics referred to in the hints below.
  1. Problem 1 will be a 5 part matching exercise, worth a total of 7.5 points (so 1.5 points for each of the 5 answers). All you have to do is match 5 descriptions of possible models to 5 sets of p-values. The setting will pertain to analysis of covariance.
  2. Problem 2 will be a 8 point problem, with 3 parts. The parts pertain to an experimental design consideration from Ch. 8 (so review blocking (especially, the purpose of blocking, and how to block effectively) and randomization), checks of relevant assumptions, and how to set things up in order to be able to test for the presence of a certain type of effect. (I recommend reviewing the procedures covered in Ch. 9 and Ch. 11 of S&W. For this problem you won't be asked to be able to perform any tests or do any numerical work. So you don't have to worry about doing computations that you're used to using SPSS to perform.)
  3. Problem 3 is a 3 point problem will request that you provide a relatively short written explanation about something to do with the size/level of a test, the power of a test, type I errors and type II errors, and p-values. (It may be that not all of these things will be considered, but I recommend that you review all of these things in order to be prepared for whatever exactly it is I request an explanation of. Something to think about: the interpretation of a smallish p-value.) I'm not going to request a definition, but will instead request an explanation to test your understanding of a particular concept.
  4. Problem 4 is a 4 point problem that will request that you provide a relatively short written explanation about something to do with the interpetation of regression coefficients (in both simple regression and multiple regression models). Again, I'm not going to request a definition, but will instead request an explanation to test your understanding of a particular concept.
  5. Problem 5 will be a 5 part matching exercise, worth a total of 7.5 points (so 1.5 points for each of the 5 answers). All you have to do is match 5 different nonparametric tests to 5 different data sets and/or descriptions of situations. The focus is on which test is possibly appropriate for which situation. (There is no need to actually do the tests to perform the matching.)