Required Homework 1

1

A sports researcher gives a random sample of Mission college students a questionnaire asking how many

minutes they exercise this week. Out of 200 Mission College students given the questionnaire, 100 filled it out

and returned it. Based on this questionnaire, the researcher concluded that the average Mission College student

exercises for 3.5 hours each week. Select one option below:

a.

This study will not suffer from any kind of bias.

b. This study will suffer from response bias, but not non-response bias.

c.

This study will suffer from non-response bias, but not from response bias.

d. This study will suffer from both response bias and non-response bias.

2

In problem 1, what is the frame: (Select one)

a.

All California adults who participate in sports.

b. All Mission College students.

c.

Everyone who cares about their health.

3

A congressional district consists of three areas which are different in economics, culture, and ethnic mix. A

political pollster wants to do a proportional stratified sample. Area 1 has 25,000 residents, Area B has 50,000

residents, and Area C has 75,000 residents. The total sample size will be 3000. How many from each area

should be in the sample?

4

Before Dr. David Ho developed the "Three Drug Cocktail" for treating AIDS (which is now the standard

treatment) the only effective treatment for AIDS was AZT. Unfortunately, AZT had limited effectiveness: it

slowed the progress of the disease, but did not provide a cure, and almost everyone treated with AZT eventually

died. When the "Three Drug Cocktail" went into clinical trials, the patients were randomly assigned to one of

two groups. Sample #1 got the new, unproved "Three Drug Cocktail" treatment, while sample #2 was treated

with AZT. The patients did not know which treatment they got, nor did the doctors who evaluated them know.

The patients were evaluated at a single point in time after treatment was completed. The sample treated with

the "Three Drug Cocktail" did so much better than the sample treated with AZT that, in accordance with US law,

the experiment was halted and all the patients were given the "Three Drug Cocktail". The "Three Drug Cocktail"

is now the standard treatment (although it has been improved since the original clinical trials) and AIDS is now in

most cases a chronic, treatable disease rather than a terminal illness.

a.

This is an observational study. TRUE FALSE

b. This study proves that the "Three Drug Cocktail" is a better treatment than AZT. TRUE FALSE

c.

In statistical terminology, what is the "treatment" in this study?

d. This is a double blind study. TRUE FALSE

e. This is a cross sectional study. TRUE FALSE

5

Using the same data as problem 3 (the "Three Drug Cocktail" clinical trials) what is the control group?

a.

The sample who took AZT.

b. The sample who took the "Three Drug Cocktail".

c.

The scientists who organized the study.

Required Homework 1

6

An educational study attempts to determine if poorly performing grade school teachers can improve if they are

mentored by an experienced teacher who is a high performer. 200 poorly performing teachers were grouped

into pairs; the two teachers in each pair were performing at the same (low) level. For each pair, one received

mentoring and the other did not. (The choice of which would be mentored was random.) Their performance is

evaluated again a year later.

This is systematic sample. TRUE FALSE

This is a matched pair design. TRUE FALSE

This is a designed experiment. TRUE FALSE

What is the treatment in this study?

What is the response variable?

7

An educational study is planned to determine whether having taking a music class reduces the chance that a

student will drop out of high school. The study is performed in three high schools with a high dropout rate.

Before the study none of the high schools had music classes (due to lack of funding). Students are randomly

assigned to either take a music class or not to take a music class. The study finds that students who take music

are less likely to drop out.

Which group of students is the control group?

8

A department store wants to get an idea of how satisfied their customers are. The store assigns an employee to

interview every 100

th

customer leaving the store. This is an example of (select one):

a.

A designed experiment (controlled experiment).

b. The use of a placebo.

c.

A systematic sample.

d. Multistage sampling.

9

In the 19

th

century, both physicians who treated humans and veterinarians had few effective remedies for their

patients- therefore, placebos were sometimes used. Would you expect that a placebo would have more effect

on a human patient, more effect on a farm animal treated by a veterinarian, or an equal effect on both? Give a

reason for your answer.

10

In the US presidential election of 1932, Democrat Franklin D. Roosevelt defeated incumbent Republican

president Herbert Hoover. Shortly thereafter, the magazine "The New Yorker" published a cartoon showing a

lady wearing a mink coat and a pearl necklace saying "I don't know how this happened, everyone I know voted

for Hoover." What was this lady's statistical error?

a.

Her friends are not a random sample- if she can afford a mink coat and pearls, her friends are probably

rich too, and rich people tend to vote Republican.

b. Obviously, her friends didn't turn out to vote- she should have made sure they actually voted.

Required Homework 1

11

An educational study was designed to determine if having a lot of extracurricular activities (sports, choir,

orchestra, theater) in a high school reduces educational achievement by distracting students from their studies.

The researchers randomly selected 100 California high schools and then within each high school randomly

selected 200 students in each high school. For each high school they listed the number if extracurricular

activities and also the score of the students on standardized tests. To their surprise, they found that schools

with lots of extracurricular activities had much higher test scores than schools with few extracurricular activities.

Is this a designed experiment or an observational study?

Does this prove that lots of extracurricular activities is a cause of high test scores?

Can you think of a lurking variable that might affect this study?

12

In the 1950s, most Americans who drank alcohol drank either beer or hard liquor. An association of California

wine growers hired an advertising consultant named Vince Packard to run a campaign to get Americans to drink

wine. Mr. Packard ran an ad campaign conveying the idea that young, good-looking, wealthy, educated,

sophisticated people drank "dry white wine". Packard's ad agency did a survey, after the ad campaign, to

determine how effective his ad campaign had been. People who were surveyed were asked what kind of

alcoholic beverage, if any, they preferred. If they replied "a dry white wine" they were asked an additional

question: "Wine is a liquid, what does it mean to say that it is 'dry'?" Amazingly, more than half the people who

said "I like a dry white wine" had no idea what "dry" meant when applied to wine.

The people who said that they like "dry white wine" while they had no idea what a "dry wine" was were an

example of (select one):

a.

A non-representative, non-random sample.

b. The use of placebos.

c.

Too small a sample.

d. A controlled experiment (designed experiment).

e. Response bias.

13

A city government keeps track of the number of parking tickets issued each day. This is (select one):

a.

Discrete data

b. Continuous data

c.

Qualitative data

14

A survey on support for capital punishment asked the question: "Do you support capital punishment for

premeditated murder? " got 55% "yes" and 45% "no". A second survey of the same population asked the

question "If a person is convicted of premeditated murder, should they get the death penalty, life imprisonment

without the possibility of parole, or life imprisonment with the possibility of parole?" The second survey got

48% for capital punishment, 40% for life imprisonment without parole, and 12% for life imprisonment with the

possibility of parole.

The difference between the two surveys is best explained by:

a.

The first survey used a placebo.

b. The result of surveys can depend on how you word the question and what alternatives are offered.

c.

This can only be explained by a biased sample.

Required Homework 1

15

The late Harvard University paleontologist and bio-statistician Stephen J. Gould used the following example in

his class. He compared the weight of his children's pet turtle to the price (per pound) of Swiss cheese. He found

that in years when the price of Swiss cheese was high, the turtle was also larger. What can we conclude from

this? (Select one.)

a.

Time is a lurking variable. Over time, the turtle gets bigger, and the price of Swiss cheese increases due

to inflation.

b. The turtle is gaining weight because he eats too much Swiss cheese. In fact, the Gould family turtle eats

so much Swiss cheese as to create a shortage, so the price of Swiss cheese increases due to supply and

demand economic factors.

c.

The turtle gets fat because he's anxious about whether his family will be able to afford to feed him Swiss

cheese, given the price increase.

16

Which of the following is qualitative data (select one):

a.

A psychologist classifies patients by personality type.

b. A website keeps track of the number of people who visit the site each day.

c.

A veterinarian keeps track of the blood sugar of a diabetic cat.

d. A typing teacher measures the typing speed of his students in words per minute.

e. Mission College keeps records of the number of students in each class.

17

A study is designed to identify whether there is a link between drinking alcohol and breast cancer. The study

involves 300 women who have been diagnosed with breast cancer and 300 women who have not been

diagnosed with cancer. Each woman was asked how many alcoholic drinks she had per week, on the average,

during their adult years. The women with cancer reported drinking significantly more than the women who did

not have cancer.

This is a controlled experiment (designed experiment). TRUE FALSE

This is a retrospective study. TRUE FALSE

This study will probably suffer from response bias. TRUE FALSE

This is an observational study. TRUE FALSE

18

A medical study is designed to test a new drug to treat pneumonia. Patients are randomly assigned to get the

new drug or an older drug which is not completely satisfactory. After 5 days of treatment, the patients are

evaluated to see how much they have improved. The patients do not know which dug they are getting but the

doctors evaluating them do. Select one option below:

(a) This is a blind study but not a double blind study.

(b) This is a longitudinal study.

(c) This is both a blind study and a double blind study.

19

An educational study wants to evaluate the effectiveness of teaching algebra online. The population of interest

is all community colleges in California. Five California community colleges are randomly selected. At each

college, half the algebra students are randomly assigned to traditional classes and half are assigned to online

classes. And the end of the semester they all took the same final exam, and the results were compared .

This is a cluster sample. TRUE FALSE

This is an an observational study. TRUE FALSE

This is a stratified sample. TRUE FALSE