The Statistical Sleuth : A Course in Methods of Data Analysis, 2nd Edition, by Ramsey and Schafer

Before Friday, August 24

Read pages 1 through 22 of Chapter 1 lightly. Section 1.3.2 is particularly important. Many concepts in this chapter may seem new to you, even if you have recently taken a statistics course. At this stage, don't worry about understanding or computing a p-value -- we'll come back to that. Read pages 1- 15 and 21-22 of Chapter 1 thoroughly. After the class you should be able to distinguish between an observational study and experiment. Read pages 15 through 22. Also read the Conceptual Exercises on pages 22 through 24.

Before Monday, August 27

Read pages 28 through 37 of Chapter 2, and don't be discouraged by all the information that is packed here; just try to understand everything, even if that means reading it several times.

Before Wednesday, September 5

• Review pages 28 through 37 of Chapter 2 again. Note that in Display 2.6 there should be an n under the square root sign in step 2.
  
• You may still wonder why we look up 0.975 in the t-table in order to compute a 95% confidence interval. A somewhat technical explanation is given here, though this explanation is not essential and may be skipped.
• Read section 2.3 on two-sample inference

Before Friday, September 6

Read Section 2.4 on randomized tests and look here for an example.

Before Monday, September 10

Read pages 56-68, paying close attention to the concepts covered in Displays 3.4 through 3.7.

Before Wednesday, September 12

Read pages 68-76. On page 70 after "(and since the log preserves the ordering)" the equation should read Median[log(Y)]=log[Median(Y)].

Before Friday, September 14

Review Wednesday's reading material.

Before Monday, September 17

Read pages 85-95 in Chapter 4.

Before Wednesday, September 19

Review Monday's reading assignment.

Before Friday, September 21

Read pages 99-105 and the Conceptual Exercises of Chapter 4 (and their solutions also).
Technical note: The first sentence of Section 4.4.1 should read ``The sign test is a quick, distribution-free test of the hypothesis that the median difference of population pairs is zero''. In other words, the sign test examines whether 50% of the differences in population pairs are above zero (and 50% of the differences in pairs are below zero). This explains why in the sign test we compare the number (called K) of pairs in the sample that have a positive difference against n/2, since we would expect half (namely n/2) of all ̉nÓ non-zero pairs in the sample to be positive if the median difference of population pairs is zero.
We may not discuss Section 4.5 in class. Sections 4.5.1 and 4.5.2 are important. It is also good to know about the material in Sections 4.5.3 and 4.5.4, but you will not be asked any questions about this material in homework or exams.

Before Wednesday, September 26

Start reading Chapter 5 (pages 113-127).

Before Monday, October 7

Read Sections 5.5, 5.6.1, 5.6.4, and 5.7. Also try Conceptual Exercises 2-4 and 7-12. Note that we are skipping a few sections including Section 5.6.2. Section 5.6.2 talks about a rank-based method that is a generalization of the rank-sum test to the case of more than two groups. I want you to know that such a method exists in case you ever need to use it in your research, but we won't work with it in 401 this semester.

Before Friday, October 12

Read pages 149-159: Sections 6.1 and 6.2 (you may skip the subsection on Comparing rates).

Before Monday, October 15

Read pages 159-169. Try to understand the reason for using the Tukey or Tukey-Kramer procedure. Don't worry about the details of carrying it out (e.g., ignore the stuff about tables and q-values). We will learn how to get Tukey and Tukey-Kramer confidence intervals and adjusted p-values using SAS. I won't expect you to know Scheffe's procedure or about any of the procedures described in the "Others" subsection.

Review Section 5.5. We will talk about diagnostics and assumptions in the lab.

Wednesday and Friday, October 17--19

Read pages 174 to 182 (sections 7.1 through 7.3.3). Concentrate on the concept of linear regression and the least-squares regression line, even if you can't completely follow the case studies.

Before Monday, October 22

Review sections 7.1 through 7.3. Read pages 186-192 (ignore section on Computing Centering Trick). Read section 7.5.

Before Friday, October 26

Read sections 8.1 through 8.4

Before Monday, October 29

Read sections 8.5.2 and 8.6.1

Before Wednesday, October 31

Read sections 8.5 and 8.6 (try to understand the "lack of fit" principle)

Before Friday, November 2

Read sections 9.1, 9.2, 10.1, 10.2. Note that we will always use some software to compute estimates of regression coefficients and compute standard errors of these estimated regression coefficients.

Before Monday, November 5

Read section 9.3.

End of material covered by Exam 2. Start preparing for the exam.