Assignment for Monday, April 10 (M 358K, Sp 06, Smith)

• The discussion on pp. 555 - 556 of why our estimator of the difference of the population proportions is an unbiased estimator, and why the formula for the standard deviation of the difference in proportions is correct.
• Why the confidence interval is only approximate.
• The conditions (bottom of box on p. 556) on when this procedure is appropriate.
• Note that there is an error in the second to last column of the Crunch It! output in Figure 8.4 -- the lower limit of the confidence interval is wrong. (You can figure out how the error was made by looking at Figure 8.5 on p. 564.)
  
• How to modify the first procedure to use the Plus Four method (pp. 558 - 560), and when the Plus Four method is appropriate (bottom of box on p. 559).
• The caution on p. 560.
• The reasoning behind the "pooled standard deviation" for a hypothesis test for the difference of two proportions (pp. 561 - 562).
• The rule of thumb for when the significance test for comparing two proportions is appropriate. (Sentence below box on p. 562).
• The concept of relative risk. (pp. 563 - 565).