AP Statistics: Survey Methods
We wish to choose a simple random sample of size three from the following employees of a small company. To do this, we will use the numerical labels attached to the names below. 1. Bechhofer 4. Kesten 7. Taylor 2. Brown 5. Kiefer 8. Wald 3. Ito 6. Spitzer 9. Weiss We will also use the following list of random digits, reading the list from left to right, starting at the beginning of the list. 11793 20495 05907 11384 44982 20751 27498 12009 45287 71753 98236 66419 84533
c. Bechhofer, Taylor, Weiss.
Use Scenario 4-3. Which of the following statements is true?
If we use another list of random digits to select the sample, the result obtained with the list actually used would be just as likely to be selected as any other set of three names.
When controlled experiments are impractical or unethical, which of the following would be necessary to establish a cause-and-effect relation between two variables?
strong association between the variables, an association between the variables is observed in many different settings, the alleged cause is plausible, there is no other obvious variable whose effect is confounded with the explanatory variable in the study (all of the above)
Use Scenario 4-1. The sample for the survey is
the 20 people who gave the sportswriter their opinion
A farmer wishes to determine which of two brands of baby pig pellets, Kent or Moormans, produces better weight gains. Two of his sows each give birth to litters of 10 pigs on the same day, so he decides to give the baby pigs in litter A only Kent pellets, while the pigs in litter B will get only Moormans pellets. After four weeks, the average weight gain for pigs in litter A is greater than the average weight gain for pigs in litter B. Use Scenario 4-7. The brand of pellets is
the explanatory variable
Just before the presidential election of 1936, the magazine Literary Digest predicted—incorrectly, as it turned out—that Alf Landon would defeat Franklin Delano Roosevelt. Landon lost in a landslide. It turned out that the magazine had only polled its own subscribers, plus others from a list of automobile owners and a list of people who had telephone service. All three groups had higher than typical incomes during the Great Depression. This is an example of
undercoverage
Use Scenarion 4-8. Which of the following best describes the inferences the researchers can make based in his results?
They can make inferences about cause and effect, but not about the populations from which the samples were taken
A news release for a diet products company reports: "There's good news for the 65 million Americans currently on a diet." Its study showed that people who lose weight can keep it off. The sample was twenty graduates of the company's program who endorse it in commercials. The results of the study are probably
biased, overstating the effectiveness of the diet
A researcher observes that, on average, the number of traffic violations in cities with Major League Baseball teams is larger than in cities without Major League Baseball teams. The most plausible explanation for this obsverved association is that the
Association is due to the presence of a lurking variable (Major League teams tend to be in large cities with more people, hence a greater number of traffic incident)
Use Scenario 4-4. Another correct choice of labels for the 816 students is
Both (B) and (C) are correct.
A large suburban school wants to assess student attitudes towards their mathermatics textbook. The administration randomly selects 15 mathematics classes and gives the survey to every student in the class. This is an example of a
Cluster sample
Simple random sampling
DOES NOT: reduce bias resulting from poorly worded questions, offset bias resulting from under coverage and nonresonse, reduce bias resulting from the behavior of the interviewer or reduce variability (none of the above)
Use Scenario 4-5. Which of the following best describes the inferences the researcher can make based on his results?
He can make inferences about the population from which the samples were taken, but not about cause in effect
Which of these statements about the table of random digits is true?
None of these is true.
To determine employee satisfaction at a large company, the management selects an SRS of 200 workers from the marketing department and a separate SRS of 50 workers from the sales department. This kind of sample is called a
Stratified random sample
You want to know the opinions of American school teachers about establishing a national test for high school graduation. You obtain a list of the members of the National Education Association (the largest teachers' union) and mail a questionnaire to 2500 teachers chosen at random from this list. In all 1347 teachers return the questionnaire. ____ 4. Use Scenario 4-2. The population is
all American school teachers
A sportswriter wants to know how strongly Lafayette residents support the local minor league baseball team, the Lafayette Leopards. She stands outside the stadium before a game and interviews the first 20 people who enter the stadium. ____ 1. Use Scenario 4-1. The intended population for this survey is
all residents of Lafayette
To determine the proportion of each color of Peanut Butter M&M, you buy 10 1.69 ounce packages and count how many there are of each color. This is an example of
cluster sampling
The principle reason for replication in designing experiments is that it
reduces sampling variability
A public opinion poll in Ohio wants to determine whether or not registered voters in the state approve of a measure to ban smoking in all public areas. They select a simple random sample of fifty registered voters from each county in the state and ask whether they approve or disapprove of the measure. This is an example of a
stratified random sample
Use Scenario 4-6. Instead of giving all students in the 8:30 section two cups of coffee, students in the 8:30 section are randomly assigned to a treatment group (two cups of coffee) or a control group (two cups of decaffeinated coffee). The coffee is so bad that students cannot tell whether they are in the treatment or the control group. As it turns out, students in both groups do better on the exam than students in the 9:30 section, who weren't given anything. This could be the result of
the placebo effect
Researchers wish to determine if a new experimental medication will reduce the symptoms of allergy sufferers without the side effect of drowsiness. To investigate this question, the researchers randomly assigned 100 adult volunteers who suffer from allergies to two groups. They gave the new medication to the subjects in one group and an existing medication to the subjects in the other group. Forty-four percent of those in the treatment group and 28% of those in the control group reported a significant reduction in their allergy symptoms without any drowsiness. Use Scenarion 4-8. The experimental units are the
100 adult volunteers
An article in the student newspaper of a large university had the headline "A's swapped for evaluations?" Results showed that higher grades directly corresponded to a more positive evalutation. Which of the following would be a valid conclusion to draw from the study?
Teacher of courses in which the mean grade is higher apparently tend to have above-average teaching evaluations
A stratified random sample is appropriate when
The population can be easily subdivided into groups according to some categorical variable, and the variable you are measuring is very similar within the groups but quite different between groups.
Eighty volunteers who currently use a certain brand of medication to reduce blood pressure are recruited to try a new medication. The volunteers are randomly assigned to one of two groups. Blood pressure is measured before, during, and after the study. Which of the following best describes a conclusion that can be drawn from this study?
We can determine whether the new drug reduces blood pressure more than the old drug for individuals like the subjects in the study
A stratified random sample addresses the same issues as which of the following experimental designs?
a block design
A study sponsored by American Express Co. and the French government tourist office found that old stereotypes about French unfriendliness were not true. The respondents were more than 1000 Americans who have visited France more than once for pleasure over the past two years. The results of this study are probably
biased, overstating the extent to which the old stereotypes were not true
The reason that blocking (as in a randomized block design) is sometimes used in experimentation is to
compensate for anticipated differences in the response variable for different values of a specified "blocking" variable
The principle reason for the use of random assignment in designing experiments is that it
creates approximately equal groups for comparison
Are dogs better at tracking the movements of white objects or red objects? Fifteen experienced "disk dogs" who have been trained to catch flying disks in mid-air are given the chance to catch a bright red disk or a plain white disk. Each disk is thrown 10 times for each dog, with the sequence of disks (red or white) determined randomly. The proportion of red disks caught to the proportion of white disks caught is compared for each dog. This is an example of a
matched pairs design
An opinion research firm wants to find the country's reaction to a speech by a famous politician. They randomly select six states, then randomly select ten Zip Codes from each state. Fifty people from each Zip Code are randomly selected for the survey. This is an example of
multistage sampling
In order to select a sample of undergraduate students in the United States, I select a simple random sample of four states. From each of these states, I select a simple random sample of two colleges or universities. Finally, from each of these eight colleges or universities, I select a simple random sample of 20 undergraduates. My final sample consists of 160 undergraduates. This is an example of
multistage sampling
Use Scenario 4-6. Suppose half of the 8:30 students are randomly allocated to the treatment group (two cups of coffee), the other half to the control group (two cups of decaf). In addition, half of the 9:30 students are randomly allocated to the treatment group, the other half to the control group. This is an example of a
randomized block design
In an experiment, an observed effect so large that it would rarely occur by chance is called
statistically significant
A public opinion poll in Ohio wants to determine whether or not registered voters in the state approve of a measure to ban smoking in all public areas. They select a simple random sample of fifty registered voters from each county in the state and ask whether they approve or disapprove of the measure. This is an example of a
stratified random sample
You plan to give a math achievement test to samples of 15 year-olds from both the U.S. and Korea in order to compare mathematics knowledge in the two countries. In each country, you will randomly choose 300 students from low-income families 400 students from middle-income families 200 students from high-income families The sample from Korea is a
stratified random sample
You plan to give a math achievement test to samples of 15 year-olds from both the U.S. and Korea in order to compare mathematics knowledge in the two countries. In each country, you will randomly choose 300 students from low-income families 400 students from middle-income families 200 students from high-income families The sample from Korea is a
stratified random sample
We divide the class into two groups: first year students and others. We then take random samples from each group. This is an example of
stratified random sampling
Medical researchers are excited about a new cancer treatment that destroys tumors by cutting off their blood supply. To date, the treatment has only been tried on mice, but in mice it has been nearly 100% effective in eradicating tumors and appears to have no side effects. As evidence of the effectiveness of the new treatment in treating cancer in humans, these studies
suffer from lack of realism
Use Scenario 4-2. The sample is
the 1347 teachers who mail back the questionnaire
In comparative trials in medicine, the placebo effect and subconscious bias on the part of the physicians evaluating treatment outcomes can be avoided by using
the double-blind technique
A marketing research firm wishes to determine if the adult men in Laramie, Wyoming, would be interested in a new upscale men's clothing store. From a list of all residential addresses in Laramie, the firm selects a simple random sample of 100 and mails a brief questionnaire to each. The chance that all 100 homes in a particular neighborhood in Laramie end up being the sample of residential addresses selected is
the same as for any other set of 100 residential addresses
In the late 1990's Scotland was considering independence from England. An opinion poll showed that 51% of Scots favored "independence." Another poll taken at the same time showed that only 34% favored being "separate" from England. The reason these results differ by so much is that
the wording of questions has a big effect on poll results
A simple random sample is
a sample that gives every possible sample of the same sized the same chance to be selected
Use Scenario 4-7. The weight of the pigs after four weeks is
the response variable
A poll conducted by the student newspaper asked, "Who do you believe will win the Ohio State Undergraduate Student Government elections?" In order to vote, one had to access the student newspaper's Web site and record one's vote. The results of the poll were summarized in a graphic similar to the following. Which of the following statements is true about these results?
the results of the survey are unreliable because response to the survey was voluntary
In order to assess the opinion of students at the University of Minnesota on campus snow removal, a reporter for the student newspaper interviews the first 12 students he meets who are willing to express their opinion. The method of sampling used is
a convenience sample
The principle reason for the use of controls in designing experiments is that it
distinguishes a treatment effect from the effects of confounding variables
A farmer wishes to determine which of two brands of baby pig pellets, Kent or Moormans, produces better weight gains. Two of his sows each give birth to litters of 10 pigs on the same day, so he decides to give the baby pigs in litter A only Kent pellets, while the pigs in litter B will get only Moormans pellets. After four weeks, the average weight gain for pigs in litter A is greater than the average weight gain for pigs in litter B. Use Scenario 4-7. The farmer has conducted a(n)
experiment, but not a completely randomized experiement
Frequently, telephone poll-takers call near dinner time—between 6 pm and 7 pm—because most people are at home them. This is an effort to avoid
nonresponse
Use Scenario 4-7. If the farmer had fed Kent pellets to an SRS of 5 pigs from litter A and an SRS of 5 pigs from litter B, with the remaining 10 pigs getting Moormans pellets, then he would have been using
a block design
A television station is interested in predicting whether voters in its viewing area are in favor of offshore drilling. It asks its viewers to phone in and indicate whether they support/are in favor of or are opposed to this practice. Of the 2241 viewers who phoned in, 1574 (70%) were opposed to offshore drilling. The viewers who phoned in are
a voluntary response sample
Does caffeine improve exam performance? Suppose all students in the 8:30 section of a course are given a "treatment" (two cups of coffee) and all students in the 9:30 section are not permitted to have any caffeine before a mid-term exam. Use Scenario 4-6. Unfortunately, any systematic difference between the two sections on the exam might be due to the fact that the 8:30 and 9:30 classes have different instructors. This is an example of
confounding
Does caffeine improve exam performance? Suppose all students in the 8:30 section of a course are given a "treatment" (two cups of coffee) and all students in the 9:30 section are not permitted to have any caffeine before a mid-term exam. Use Scenario 4-6. Unfortunately, any systematic difference between the two sections on the exam might be due to the fact that the 8:30 and 9:30 classes have different instructors. This is an example of
confounding
An opinion research firm wants to find the country's reaction to a speech by a famous politician. They randomly select six states, then randomly select ten Zip Codes from each state. Fifty people from each Zip Code are randomly selected for the survey. This is an example of
multistage sampling
One hundred volunteers who suffer from severe depression are available for a study. Fifty are selected at random and are given a new drug that is thought to be particularly effective in treating severe depression. The other fifty are given an existing drug for treating severe depression. A psychiatrist evaluates the symptoms of all volunteers after four weeks in order to determine if there has been substantial improvement in the severity of the depression. The study would be double blind if
neither the volunteers nor the psychiatrist knew which treatment any person had recieved
Use Scenario 4-6. Instead of giving all students in the 8:30 section two cups of coffee, students in the 8:30 section are randomly assigned to a treatment group (two cups of coffee) or a control group (two cups of decaffeinated coffee). The coffee is so bad that students cannot tell whether they are in the treatment or the control group. As it turns out, students in both groups do better on the exam than students in the 9:30 section, who weren't given anything. This could be the result of
the placebo effect
Use Scenario 4-1. The newspaper asks you to comment on their survey of local opinion. You say:
this is a convenience sample, it will almost certainly overestimate the level of support among all Lafayette residents
A researcher is testing the effect of a new fertilizer on crop growth. He marks 20 plants in a field, splits the plots in half, and randomly assigns the new fertilizer to one half of the plot and the old fertilizer to the other half. After 4 weeks, he measures the crop yield and compares the effect of the two fertilizers. This design is an example
Matched pairs design
100 volunteers who suffer from anxiety take part in a study. 50 are selected at random and assigned to receive a new drug that is though to be extremely effective in reducing anxiety. The other 50 are given an existing anti-anxiety drug. A doctor evaluates anxiety levels after two months of treatment to determine if there has been an larger reduction in the anxiety levels of those who take the new drug. This would be double blind if
Neither the subjects nor the doctor knew which treatment any subject had recieved
A new cough medicine was given to a group of 25 subjects who had a cough due to the common cold. 30 after taking the new medicine, 20 of the subjects reported that their cough had disappeared. From this information you conclude
Nothing, because there is no control group for comparison
A candidate for mayor of Dallas wants to know what proportion of city residents support his candidacy. His staff calls 1,000 people chosen at random from the city telephone directory; 850 of them respond. What are the population and the sample in this example?
Population: all Dallas residents, Sample: the 850 people who respond
Use Scenario 4-7. The feed they get is not the only factor affecting the rate at which pigs gain weight. Genetic differences also affect weight gain. It is likely that the pigs in litter A are genetically different from the pigs in litter B, since the two litters have different mothers. Since the farmer is only interested in determining which brand of pellets is better, the study suffers from
confounding
In order to assess the effects of exercise on reducing cholesterol, a researcher took a random sample of fifty people from a local gym who exercised regularly and another random sample of fifty people from the surrounding community who did not exercise regularly. They all reported to a clinic to have their cholesterol measured. The subjects were unaware of the purpose of the study, and the technician measuring the cholesterol was not aware of whether or not subjects exercised regularly. Use Scenario 4-5. This is a(n)
observational study
For a certain experience to you have 8 subjects, of which 4 are female and 4 are male. If a randomized clock design is used with the subjects blocked by their gender, which of the following is not a possible group of subjects. Who receive treatment A?
Atwater, Bacon, Chu, Johnson
A marine biologist wants to estimate the mean size of the barnacle Semibalanus balnoides on a stretch of rocky shoreline. To do so, he randomly selected twenty 10-cm. square plots and measured the size of every barnacle in each plot. This is an example of
cluster sampling
A marine biologist wants to estimate the mean size of the barnacle Semibalanus balnoides on a stretch of rocky shoreline. To do so, he randomly selected twenty 10-cm. square plots and measured the size of every barnacle in each plot. This is an example of
cluster sampling
A researcher is testing a company's new stain remover. He has contracted with 40 families who have agreed to test the product. He randomly assign 20 families to the group that will use the new stain remover and 20 to the group that will uses the company's current product. The most important reason from this random assignment is that
Randomization is a good way to create two groups of 20 families that are as similar as possible, except for the treatments they receive
A simple random sample of size n is defined to be
a sample of size n chosen in such a way that every set of n units in the population has an equal chance to be the sample actually selected
Twelve people who suffer from chronic fatigue syndrome volunteer to take part in an experiment to see if shark fin extract will increase one's energy level. Eight of the volunteers are men, and four are women. Half of the volunteers are to be given shark fin extract twice a day, and the other half are to be given a placebo twice a day. We wish to make sure that four men and two women are assigned to each of the treatments, so we decide to use a block design with the men forming one block and the women the other. A block design is appropriate in this experiment if
we believe men and women will respond differently to treatments
A stratified random sample is appropriate when
The population can be easily subdivided into groups according to some categorical variable, and the variable you are measuring is very similar within the groups but quite different between groups
