Please turn off cell phones, pagers, etc.

1
Please turn off cell phones, pagers, etc. The
lecture will begin shortly.
2
Lecture 36
Todays lecture will review key material from the
second half of the semester.
We will discuss questions similar to those you
are likely to find on the final exam.
3

1. Suppose that the rates of hypertension in a group
   of men and a group of women are .33 and .50,
   respectively. The odds ratio is

1. .33 / .50 .66

1. (.33/.67) / (.50/.50) .5

1. (.33/1.33) / (.50/1.50) .74

1. .33 / 1.50 .22

4
2. Suppose that the probability of a rare birth
defect is doubled if the mother smokes during
pregnancy. The odds ratio will be

1. very small

1. close to 1.0

1. approximately 2.0

1. there is no way to tell

5
3. Suppose that the Presidents approval rating
among all likely voters in the U.S. is 41, but
his approval rating in a sample of n1,000 voters
is 37. The difference between these numbers
(4) is called

1. Sampling variation

1. Sampling error

1. Standard error

1. Measurement error

6
Side effects?
4. A pharmaceutical company finds that 10 out of
100 patients who are using a drug report side
effects, compared to only 5 out of 100 who are
using a placebo. Find the relative risk.
Yes
No
Drug
10
90
Placebo
5
95

1. (10/90) / (5/95) 2.1

1. (10 95) / (90 5) 2.1

1. (100 15) / 200 7.5

1. (10/100) / (5/100) 2.0

7

1. A chisquare test for independence in a 22 table
   yields a p-value of .02. Which statement is
   incorrect?

1. The evidence against the null hypothesis is
   strong.

1. The two binary variables appear to be related.

C. Theres a 2 chance that the null hypothesis
is true.

1. The chisquare statistic is greater than 3.84.

8

1. You buy a raffle ticket that gives you a 1 in 100
   chance to win 5 and a 1 in 1,000 chance to win
   100. Your expected winnings are

1. 5 cents

1. 10 cents

1. 15 cents

1. 5 dollars

9

1. The Central Limit Theorem for a sample mean does
   not say that the sample mean

1. is approximately normally distributed

1. is, on average, equal to the population mean

C. has a standard deviation equal to the
population SD

1. has a standard deviation smaller than the
   population SD

10

1. Suppose that you flip a fair coin 100 times.
   Theres about a 95 chance that the percentage of
   you will observe is

1. between 45 and 55

1. between 40 and 60

C. between 25 and 75

1. between 48 and 52

11

1. Suppose a candidate has the support of 40 of all
   likely voters. A random sample of 100 likely
   voters is taken. The standard error of the
   sample proportion will be

1. the margin of error

1. the range of likely values

C. the square root of (.4 .6) / 100

1. between .4 1/10 and .4 1/10

12

1. A traffic safety engineer reviewed the locations
   of accidents in Pennsylvania in 2004 and
   identified the 100 most dangerous intersections.
   In 2005, we would expect the average number of
   accidents at those same 100 intersections to be

1. lower than it was in 2004, but still above average

B. about average
C. about the same as it was in 2004

1. higher than it was in 2004