Chapter 3

1. An English teacher encourages students to use a word processing program on the school's computers to compose and revise their papers. There are two word processing programs available, UWRITE and DOWNWRITE. The computer keeps track of the time spent by each student. The teacher studies the data and sees that the students who use UWRITE spend 20% less time writing their papers than the students who use DOWNWRITE. Is this study an experiment? Explain why or why not.

2. The teacher decides to carry out an experiment to determine which word processing program is faster for students to learn and use. The subjects will be the 32 students in next semester's course, none of whom have used either program before. The response variable will be the time required to prepare the first draft of a 5,000-word paper. Outline the design of a completely randomized experiment for the teacher to use.

3. The following class list gives the names of the 32 students. Use Table B beginning at line 120 to carry out the random assignment required by your design in Question 2.

Anderson Evans Jiang Raab

Biehn Finney Jones Reilly

Brinkman Foster Kamin Scaletta

Calderone Gallucci Kim Sego

Clark Gobel LaPlante Ting

Dale Hafen Metcalf Vasquez

Delos Hernandez Moss Vaughn

Everett Hockema Palmer Young

4. A food scientist is interested in determining how the conditions under which soybeans are stored affect the "cookability" of the beans when they are processed. She will store lots of beans at each of three temperatures and 2 humidity levels for 2, 8, and 16 weeks, and then process the beans and measure their cookability.

How many factors does this experiment have? Identify the factors and state how many levels each factor has. If all combinations are compared, how many different treatments are there in the experiment? What are the experimental units? What is the response variable?

5. A medical experiment will compare a new medication intended to reduce high blood pressure with a standard medication. The subjects are 200 men aged 40 to 60 with high blood pressure in Minneapolis and another 300 similar men in San Francisco. Outline a randomized block design for this experiment.

6. A member of Congress wants to know what the voters in his district think about a proposed gun control law. A staff member reports that 87 letters have been received on this subject, of which 73 oppose the law. That is 84% of the letter-writers oppose gun control legislation.

What is the population in this problem? (Be as specific as possible) The 87 letters are a sample. Explain why the sample result is almost certainly biased. Is the proportion of the population that opposes the proposed law probably greater or less than 84%?

7. Dr. Pimento wants to know what percent of American college students approve of television advertising. Students who take Psychology 101 at his school are required to serve as research subjects. Dr. Pimento chooses an SRS of 100 from the 900 students taking the course this semester. He asks, "Television as we know it would not be possible without advertising. Do you agree that having commercials on TV is a fair price to pay for being able to watch it?" Of the 100 students in the sample, 82 answer "yes."

a.) What is the population in this setting? Describe the parameter and the statistic.

b.) Despite the fact that Dr. Pimento used an SRS, at least two sources of bias are present here. What are they?

8. National opinion polls such as the Gallop Poll usually interview a sample of about 1500 people. Immediately before a presidential election, however, the size of the sample is increased to about 4000 people. What is the advantage of the larger sample?

9. A news item says, "54% of the people surveyed approved of the president's handling of his office. The margin of error of the poll result is plus or minus three percentage points." The margin of error announced by the polls covers the middle 95% of the sampling distribution of the sample proportion p. The sampling distribution of p is normal. Its mean is equal to the population percent p who approve of the president (no bias). What is the standard deviation of the sampling distribution of p in this case?