Chapter 3, Design: How Individuals Are Studied Video Solutions, Statistics : Looking at The Big Picture | Numerade (2024)

"Where You Will Live the Longest" is the title of a 2006 report on a study of mortality rates in the United States during the 1980s and 1990 s . A resident of one of the topranked counties for longevity (Clear Creek, Colorado) followed up the report with her explanation of why people in that area tend to stay lively: "It's because it's too risky not to . . . Don't look closely at the droppings in the driveway and you might not realize they belong to a nearby ferocious animal. Fail to notice the morning ice along the creek and you won't order your propane or woodsplitter in time before winter . . . ${ }^3$
a. What design (observational study, experiment, sample survey, or anecdotal evidence) was used to produce the original report?
b. What design did the Clear Creek resident use?

A 13 -member panel was convened by the National Institutes of Health in 2004 to look for consensus on causes of youth violence and ways to prevent it. Its conclusion was that "boot camps and other get-tough programs for adolescents do not prevent criminal behavior, as intended, and may make the problem even worse. ${ }^{74}$ Was their conclusion most likely based on one observational study, several observational studies, an experiment, a sample survey, or anecdotal evidence?

A newspaper article about a book entitled Wilderness Within, Wilderness Without, by Shannon Szwarc, tells how "living in a rugged outdoor environment with firm but nurturing counselors ${ }^{75}$ at a therapeutic wilderness camp transformed the author. Would a reader's conclusion that the wilderness program is effective be based on one observational study, several observational studies, an experiment, a sample survey, or anecdotal evidence?

The two preceding exercises describe, respectively, a study by the National Institutes of Health and a book by a former wilderness camp member. Both address the effectiveness of such camps.
a. What aspects of the National Institutes of Health study would make its claims persuasive?
b. What aspects of the book about a wilderness camp's benefits to the author would make its claims persuasive?

Suppose that get-tough programs as described in the previous exercises actually are beneficial, and a study erroneously concludes that they do nothing to prevent criminal behavior. Describe the potential harmful consequences of such a mistake.

Suppose that get-tough programs do not prevent criminal behavior, and a study erroneously concludes that they are beneficial. Describe the potential harmful consequences of such a mistake.

A paper by Emily Oster of Harvard University, published in the Social Science Research Network's electronic library in 2004, explored the "possibility that the witchcraft trials [between 1520 and 1770] are a large-scale example of violence and scapegoating prompted by a deterioration in economic conditions. In this case, the downturn was brought on by a decrease in temperature and resulting food shortages. [. . .] Witches became target for blame because there was an existing cultural framework that both allowed their persecution and suggested that they could control the weather. ${ }^{11}$ Would the relationship between witchcraft trials and weather be examined via a sample survey, an observational study, or an experiment?

A drug for female sexual dysfunction called Intrinsa was tested in 2004, both for efficacy and for possible dangerous side effects before being considered for approval by the Food and Drug Administration. A report of clinical trials states: "Fifty-two percent of those [women] given the drug said they experienced a 'meaningful benefit' in their sex-lives-but so did $31 \%$ of those who were given the placebo. ${ }^7$ Who controlled the values of the explanatory variable in this study: the women or the researchers?

"As Orders Soar, Concerns Over Stun Guns Grow" reported in 2004 that "Amnesty International says stun guns are being abused by police and wants more scientific study done to determine whether the devices are safe. Amnesty says at least 74 people have died in the United States and Canada in the past four years after being shocked with Tasers. The group also says officers have turned stun guns on the mentally disturbed, children, and the elderly. ${ }^{n 8}$ On the other hand, Taser company chairman Phil Smith says, "We get e-mail from police every week . . . thanking us for developing a weapon so they didn't have to shoot somebody." According to the article, "Phoenix police officers credit Tasers with helping police shootings drop by more than half and fatal shootings by $31 \%$ last year."
a. One issue involved here is whether or not Tasers can be lethal. Discuss the harmful consequences of incorrectly deeming them to be safe for use by police officers, if, in fact, they can cause fatal heart attacks. Discuss the consequences of deeming them to be unsafe if, in fact, the deaths mentioned in the article were not due to the Tasers.
b. Suppose some Tasers are to be tested to see if the voltage delivered is dangerously high. If a very small sample of Tasers is tested, is there more risk of concluding Tasers in general are safe when, in fact, some are not, or more risk of concluding Tasers in general are not safe when, in fact, they are?

An article entitled "Stress Found in Returning Soldiers" includes the following information: "A desperate woman wearing a black robe and her young son walked up to Alan Lewis, a United States soldier who was working security. The boy had just picked up an unexploded piece of ordnance, which blew up, badly burning his face and hands . . .
Weeks later, Lewis would lose parts of both his legs after a land mine exploded under the Humvee he was driving. But his most horrible experience, Lewis said, was the sight of the boy and his mother. ${ }^{m 9}$ The article also found that of 6,200 members of four United States combat infantry units deployed in Iraq and Afghanistan, about $75 \%$ "reported encountering injured or ill women or children who they were unable to help." It was noted that "before the Sept. 11 terrorist attacks, an estimated $8 \%$ of Americans suffered from the [post traumatic stress] disorder."
This article includes information gathered via three designs: observational study, sample survey, and anecdotal evidence.
a. Which type of design accompanies the first quote above?
b. Which type of design accompanies the second quote above?
c. Which type of design accompanies the third quote above?

For which of these is it more important to include a response option like "don't know" or "not sure": a question about whether the respondent owns a gun, or a question about whether the respondent feels the national firearms registry program should be abandoned?

A 2002 survey of 5,000 households found that $87 \%$ of the respondents said their employers prohibited smoking in the workplace. If the survey did not permit a response of "don't know," who would be more likely to give an incorrect response about whether smoking was permitted at their workplace-smokers or nonsmokers?

For which question is it more important to include an option of "don't know": asking if respondents have a permanent tattoo on their body, or asking how many people will be present at their upcoming Thanksgiving dinner?

Many questionnaires use a scale of 1 to 5 for responses.
a. Would such questions be considered open or closed?
b. Explain why a scale of 1 to 5 is much more common than a scale of 1 to 4 or 1 to 6 .

An article entitled "Survey Finds County Enjoys Good Health" reports that "the vast majority of [county] residents rate their health as good to excellent . . . $20 \%$ of nearly 5,000 adults surveyed described their health as excellent, $34 \%$ called it very good, and $30 \%$ called it good. Sixteen percent said they were in fair or poor health. ${ }^{12}$
a. If the questionnaire included five response options, what were they?
b. If the questionnaire included four response options, what were they?
c. In which case would you say the options were unbalanced: if there were five options given, or four, or both of these, or neither?

In the year 2000, Hamilton College sponsored a national survey to probe high school students' attitudes toward gun issues.
The survey included a question about whether or not the respondents were in favor of the following: "Require a 5-day waiting period between the purchase and delivery of a handgun, to keep the buyer from acting on impulse against himself or others. ${ }^{13}$ Did the surveyors want respondents to be in favor or opposed; or is it impossible to tell from the wording of the question?

The following survey questions were posted on the website www.a-human-right.com:
1. Is it morally correct tor a police otticer to shoot in order to save an innocent victim from an attack?
2. If no police officer is present is it morally correct for the innocent victim to shoot to protect self or dependents from an attack?
3. If my child or my spouse were assaulted, I would . . .
- Run away and hope my kid or spouse can keep up with me
- Be a good witness so I can tell the cops what happened later
- Try to convince the attacker to stop through verbal persuasion
- Fight to stop the attack
4. Given a choice, I would prefer to defend myself with . .
- My bare hands
- An ineffective weapon, such as pepper spray
- An effective weapon, such as a firearm
For which question(s) is it clear what answer the surveyors want respondents to give?

"Data Found Lacking on Effects of Gun Control Efforts" explains that an analysis in 2004 of "efforts to control violence by restricting guns says there is not enough evidence to reach valid conclusions about their effectiveness. [. . .] A serious limit in such analyses is the lack of good data on who owns firearms and on individual encounters with violence, according to the study. Research scientists need appropriate access to federal and state data on gun use, manufacturing and sales, the study urged. [...] The report calls for the development of a National Violent Death Reporting System and a National Incident-Rased Reporting System for collecting data." ${ }^{14}$
Are these scientists mainly concerned with data production, displaying and summarizing data, probability, or statistical inference?

An exit poll surveyed 11,027 voters after the 2004 presidential election. Among the questions asked were, "Which one issue mattered most in deciding how you voted for president?" 15 and "How do you feel about the United States decision to go to war in Iraq?" As far as the ordering of questions is concerned, do you think there is less chance of bias if they are ordered as above, or in reverse order, or doesn't it matter?

An exit poll surveyed 11,027 voters after the 2004 presidential election. The survey asked voters, "Compared to four years ago, is your family's financial situation about the same, better today, or worse today? ${ }^{16}$ and also asked voters to tell their income bracket for the previous year, with choices ranging from under $\$ 15,000$ to $\$ 200,000$ or more. As far as the ordering of this pair of questions is concerned, do you think it is better to order them as above or in reverse order? Explain your reasons.

Environics Research Group in Canada conducted a survey about firearms in which some respondents were asked to tell whether they agreed with the first or the second opinion described in the following: "The Auditor-General has found significant cost overruns in the national firearms registry program. Some people say that these cost overruns prove that the concept of the registry is unworkable and should be abandoned. Other people say that despite the cost overruns, the national firearms registry is still a good concept and should be completed. ${ }^{17}$ Other respondents were presented with the same pair of opinions, but in the reverse order.
a. Why were two different forms of the question asked?
b. Describe how researchers could explore whether or not the ordering of the two opinions may affect people's responses.
c. Would the survey's results tell you more about how all Canadians feel regarding firearms registry if there were 10 respondents or 1,000 respondents; or doesn't it matter?
d. Which one of the following would most improve the above survey question: make it less complicated, give respondents more of a feeling of anonymity, or make it an open question?

Students were asked to rate their instructor's preparation for class as being excellent, good, or needs improvement.
a. If a student feels the instructor does an average job, what response will he or she probably select?
b. The options provided result in biased responses to the instructor's survey; will they tend to overestimate or underestimate the quality of the instructor's preparation?
c. Should the instructor summarize results with means or with proportions?

A survey of students found that only $10 \%$ of them liked brussels sprouts.
a. Was the survey question more likely to have been open or closed?
b. Which would do more to convince you that the percentage of all students who like brussels sprouts is close to $10 \%$ : if the survey included 200 students or 20 students?

For which survey question is it more important to assure respondents of their anonymity: when asking how they rate their instructor's preparedness or when asking if they like brussels sprouts?

A survey asked, "On how many occasions (if any) have you sniffed glue, breathed the contents of an aerosol spray can, or inhaled other gases or sprays, to get high during the past 30 days?" Response options were "0 occasions, 1-2 occasions, 3-5 occasions, 6-9 occasions, $10-19$ occasions, 20-30 occasions, 40 or more occasions." An alternative wording would have been "Have you sniffed glue, breathed the contents of an aerosol can, or inhaled other gases or sprays, to get high during the past 30 days? If so, on how many occasions? $\qquad$ " For which wording are students more likely to admit (or simply claim) to have engaged in the above-mentioned behaviors, the first or the second?

Phantom Illness" poses the question, "Do you ever call in sick when you're not sick?" and reports results of a poll by CareerBuilder.com: "More than a third of workers, or $35 \%$, said they had done so at least once in the past year." ${ }^{18}$ Tell why there may be less bias in the responses to this e-mail survey than if it had been conducted by phone or in person.

If students are asked by their statistics instructor in the course of the semester to complete a survey about quality of instruction, which method is more likely to elicit honest responses: asking them to complete the survey via computer lab facilities or via a pencil-and-paper survey?

A survey asked, "How many close friends do you have?" Which of these is most worrisome: leading question, complicated question, sensitive question, or hard-todefine concept?

For each of these studies, a precise summary is used to assess an imprecise thing. Tell whether the imprecision involves the explanatory variable or the response variable.
a. A study in which people who were asked to think about bad hair scored 3 points lower on self-esteem than people who were asked to think about leaky packages.
b. A study that showed happy people lived 20 months longer than unhappy people.

"Moderate Walking Helps the Mind Stay Sharper" is the title of an article that reports, "Women who walked $11 / 2$ hours per week had a $20 \%$ lower risk of mental impairment than those who walked less than 40 minutes per week, according to a study of 18,766 elderly women conducted by researchers from the Harvard School of Public Health." ${ }^{19}$
a. Was this an observational study or an experiment?
b. Which one of these would be most worrisome here: confounding variables, people's behaviors affected by their awareness of participating in a prospective study, or anecdotal evidence?
c. Suppose age was thought to be a possible confounding variable in the relationship between walking and mental fitness. How could researchers handle this problem?
d. Report the explanatory variable using a single word from the article's title $\qquad$ and identify its variable type:
1. Categorical
2. Quantitative
e. Report the response variable using two words from the article's explanation in
quotations $\qquad$ and identify its variable type:
1. Categorical
2. Quantitative
f. Do we compare means or proportions?

An article entitled "Men, Beer, and Lung Cancer" reported that men who regularly consumed red wine were less likely to contract lung cancer than were men who drank beer.
a. The fact that men were studied separately from women indicates that a certain variable is being controlled for as a potential confounding variable. What is it?
b. Another variable was not mentioned and could easily enter in as a confounding variable. What is it?

A study compared percentages of men and of women who were illiterate in various countries around the world. For instance, in Zimbabwe, $7.2 \%$ of the men and $15.4 \%$ of the women were illiterate. Is this a paired or a two-sample study?

Pell grants provide federally funded financial aid to low-income students. A study looked at the percentage of students with Pell grants in 1992, compared to 2001, for the nation's top 20 universities. For instance, Harvard went from $4.6 \%$ of its students having Pell grants in 1992 to $6.8 \%$ in 2001 , and Massachusetts Institute of Technology went from $16.3 \%$ to $12.4 \%$.
a. Was this a two-sample or paired study?
b. Tell which of the following processes would not be relevant in this situation, and why: data production, displaying and summarizing data, or statistical inference.

"TV Pro Wrestling Spurs Teen Violence" is the headline of an article that reports on a study that related the viewing of pro wrestling on television with violent behavior such as date fights (verbal or physical abuse taking place in social settings traditionally known as "dates").
a. Describe how data could have been gathered via a retrospective observational study; what is a major drawback to such a study design?
b. Describe how data could have been gathered via a prospective observational study; what is a major drawback to such a study design?
c. Assuming there is, in fact, some relationship between watching pro wrestling and teen violence, give a brief alternative explanation that is different from the headline's claim that "TV pro wrestling spurs teen violence."

An article entitled Couch potato nation describes a report by the federal Centers for Disease Control and Prevention that "analyzed responses from 119,000 people who were asked in 1996 about their physical activity during the month prior to the survey. ${ }^{20}$
a. This was an observational study. Was it retrospective or prospective?
b. The article states that overall, nearly $30 \%$ of adults reported being inactive. Is physical activity treated here as a quantitative or a categorical variable?
c. It was also reported that the rate of inactivity was $37 \%$ in small towns, compared with a rate of $27 \%$ in large urban areas. Now there are two variables of interest; describe the explanatory variable and tell whether it is quantitative or categorical.

A study was carried out to see whether acupuncture can relieve stress; another study looked at whether Rogaine results in significant hair restoration.
a. Which of the two studies is more susceptible to the placebo effect? Explain.
b. For which of the studies would it be easier to make the subjects blind to whether or not they are receiving the actual treatment?

A news article reports on a study published in Lancet in 1996: "Researchers experimented on themselves, letting honeybees sting them repeatedly and then comparing whether scraping a stinger after two seconds resulted in a smaller weal than yanking it out as quickly." ${ }^{12}$ Why did the researchers compare weal sizes, instead of simply reporting how painful the sting was?

In September of 2006, Scientific American reported that "Washing Hands Reduces Moral Taint. ${ }^{23}$ To demonstrate that people are inclined to wash their hands as a response to a feeling of moral uncleanliness, researchers asked some students to recall an ethical action they had undertaken in their lives; others were asked to recall an unethical action. "After recalling an unethical memory, students were more likely to choose a free antiseptic wipe over a free pencil when offered the choice, as compared with untested subjects who showed little preference." Which of these would be most worrisome in this experiment: placebo effect, experimenter effect, lack of realism, or imposing a treatment that is unethical?

In which study would noncompliance be more of a problem: a study where participants are to think about bad hair or leaky containers, or a study that asks participants to have tissue removed from their upper arm?

"Written Word Helps Wounds Heal" describes a British study: "In the study, which involved 36 people, half were asked to write about the most upsetting experience they had had, spelling out how they had felt. The rest of the study participants wrote about trivial things, such as how they spent their free time. Both groups spent 20 minutes a day for three days writing. Following the writing exercise, researchers created a small skin puncture [under anesthesia] on the participants' upper arms. The wounds were examined two weeks later. It was found that the group who had written about their emotional experiences had smaller wounds, meaning they had healed more quickly. Those whose wounds were healing more slowly were found to have higher levels of stress and psychological distress. "24
a. Explain why this experiment used volunteers, rather than a random sample.
b. Explain why some participants were asked to write about how they spent their free time.
c. Describe the response variable and explain whether it was likely to have been treated as quantitative or categorical.
d. Would the study's results be summarized with means or proportions?
e. Which of these is most problematic in attempting to use the study's results to conclude that writing about disturbing events can have healing benefits: placebo effect, experimenter effect, lack of realism, or imposing a treatment that is unethical?
f. What modification to the experiment's design would have been appropriate if researchers suspected that the benefits of releasing emotions by writing differs for men and women?
g. If researchers had recruited married couples and randomly assigned the man or woman to one or the other treatment, what would the design be called?

An article entitled "Study Offers Surprise on Working of Body's Clock" tells about a randomized controlled double-blind study in which subjects spent several days in a sleep lab. At night they "stayed awake in a dimly lighted room, reclining in a chair with a table over their laps. A thick black material was draped over their legs and fastened to their waists. Underneath this skirt, a knee pad with a fiber optic tube was attached to the back of their knees and a bright light was delivered through the tube for three hours. ${ }^{25}$ Other subjects were kept under identical conditions except the light was not turned on. Those treated with light had significant changes in their body rhythms, while those untreated did not.
a. Which of these would be most worrisome in this experiment: placebo effect, experimenter effect, lack of realism, or imposing a treatment that is unethical?
b. Was the black skirt used to offset the placebo effect, the experimenter effect, lack of realism, or imposing a treatment that is unethical?
c. If body rhythms are assessed by measuring body temperature and level of melatonin in the blood, is it important for researchers to be unaware of which subjects were in the treatment group when it comes time to evaluate their responses?

Keeping in mind that volunteers in experiments must almost always be paid for their time, would you expect the study discussed in Exercise 3.41 about the effect of light on humans' body clocks to have been carried out on a fairly small or fairly large sample of subjects?

Which type of studies do you think tend to use larger sample sizes-experiments or observational studies-or do you think type of study plays no role in sample size used?

For which of these studies would it be easier to include a placebo so that subjects are blind to treatment: an experiment to test if piano lessons improve math scores, or an experiment to test if taking a blood pressure medication reduces the risk of heart attacks?

An article entitled "Piano Lessons Boost Math Scores" states that "second-grade students who took piano lessons for four months scored significantly higher on math than children who did not. ${ }^{n 26}$
a. Tell why these results are more convincing if they are based on an experiment than they would be if they were based on an observational study.
b. One possible design is to take a sample of second graders and randomly assign somè tô bé givèn pianoò lessóns for 4 months, and others not. Is this a two-sample or a paired design?
c. Another possible design is to take a sample of second-grade twins and randomly assign one twin to be given piano lessons for 4 months, the other not. Is this a two-sample or a paired design?
d. Explain why the design described in part (c) could do a better job at pinpointing the effects of piano lessons on math skills, compared to the design described in part (b).
e. If an experiment is carried out to determine if piano lessons can increase math scores, is it important for the researchers who evaluate the response to be unaware of whether or not a student was in the treatment group?

Which could be called a before-and-after design: a study to see if people can lower their cholesterol by eating oatmeal every day for a month, or a study to see if breast-fed infants have fewer infections than do bottle-fed infants?

Describe how a study could test the effectiveness of the hair restoration drug Rogaine using a combination of paired and two-sample designs.

The article "When Your Hair's a Real Mess, Your Self-Esteem Is Much Less" describes a study wherein participants "were divided into three groups. One group was questioned about times in their lives when they had bad hair. The second group was told to think about bad product packaging, like leaky containers, to get them in a negative mindset. The third group was not asked to think about anything negative. All three groups then underwent basic psychological tests of self-esteem and selfjudgment. The people who pondered their bad-hair days showed lower self-esteem than those who thought about something else. ${ }^{27}$
a. Was this an observational study, an experiment, or a sample survey?

An article entitled "How Healthy Is Our State for Kids? Study Will Tell" describes a study that would recruit and enroll participants in 96 locations over the next 4 years and track them from conception through their children's twenty-first year. Participants would have at least 15 inperson visits by local research teams. For the $$\$ 2.7$$ billion study, "workers will collect biological samples from the mother and child, as well as air, water, dirt and dust from the child's environment. The study also will examine many aspects of children's lives: genetics, neighborhoods, schools, chemical exposures linked to the atmosphere, food or water supplies, and social and behavioral environments." ${ }^{28}$
a. Will this be an observational study or an experiment?
b. Which of these would be most worrisome here: treatments that are unethical to impose, people's behaviors affected by their awareness of participating in a prospective study, or people's faulty memories in a retrospective study?

A 2004 article entitled "Study of Pesticides and Children Stirs Protest" describes a proposed Environmental Protection Agency (EPA) study to see how small children absorb pesticides and other household chemicals. The EPA would pay each family " $\$ 970$, some children's clothing and a camcorder ${ }^{29}$ in exchange for their participation for two years, during which they would use various household chemicals and have their children tested for absorption rates. [Note: In 2005 it was decided not to carry out the study.]
a. Would this be an observational study or an experiment?
b. Which of these would be most worrisome here: treatments that are unethical to impose, people's behaviors affected by their awareness of participating in a prospective study, or people's faulty memories in a retrospective study?

Suppose a certain type of pesticide is actually not toxic to humans, but a study erroneously concludes that it is toxic, and the pesticide is banned.
a. Describe the potential harmful consequences of such a mistake.
b. Tell which group would be more in favor of avoiding this type of mistake: pesticide manufacturers or environmentalists.

Suppose a certain type of pesticide actually is toxic to humans, but a study erroneously concludes that it is not toxic, and the pesticide is kept in use.
a. Describe the potential harmful consequences of such a mistake.
b. Tell which group would be more in favor of avoiding this type of mistake: pesticide manufacturers or environmentalists.

Suppose that the effect of pesticides on children's absorption levels is thought to depend to some extent on two additional variables, namely the extent of previous exposure to those pesticides and the age of the child. For which of these two variables would it be easier to separate the data when examining the study's results?

One possible design to study the benefits to infants of breast milk would be a very large observational study. Another possibility would be a blind experiment wherein a very small group of women are provided daily with bottles of breast milk to feed their infants, while another small group is provided with bottles of formula. If differences between breast-fed and bottle-fed infants are, in fact, rather subtle, which study would be more likely to fail to find clear evidence of breast milk's benefits?

In 2003, The National Research Council estimated that obesity occurs in $25 \%$ of dogs and cats in Westernized societies. (Obesity can be determined by squeezing an animal in the lower chest area and seeing if the ribs can be felt.) Perhaps the $25 \%$ estimate was obtained by calling homeowners and, if they had a cat or dog, asking them to see if the animal's ribs could be felt by squeezing. Alternatively, the estimate could have been obtained by asking veterinarians to report what percentage of the animals they treat could be classified as obese, using the above method.
Keeping in mind that bias may arise either from a nonrepresentative sample or from inaccurately assessing a variable,
a. Which approach is more likely to lead to a nonrepresentative sample: calling homeowners or asking veterinarians?
b. Which approach is more likely to lead to inaccurate assessments: calling homeowners or asking veterinarians?

An article entitled "Family Dinners Benefit Teens" claims that "eating dinner together as a family is one way of keeping teenagers well adjusted and out of trouble. ${ }^{32}$
a. The article's title suggests that which variable is explanatory and which is response?
b. Compose a brief title that would suggest the opposite assignment of explanatory and response roles.
c. The article goes on to say that "adjusted teens . . . ate with their families an average of five days a week; nonadjusted teens ate with their families only three days a week." This statement implies that eating dinner with the family is the response variable; is it being treated as quantitative or categorical?

Tell which process-data production, displaying and summarizing data, probability, or statistical inference-is the focus of each of these details concerning a survey about famous athletes:
a. The website includes a pie chart showing survey results, and reports that the highest percentage ( $38 \%$ ) was for respondents who felt Michael Jordan was the most famous athlete.
b. The website suggests we can conclude that the percentage of all Americans who consider Michael Jordan to be the most
famous athlete is $38 \%$, with a margin of error equal to about $3 \%$.
c. The website tells who was included in the survey and how the question was phrased.

A survey of 122 families with epileptic children explored the behavior of the family dog in connection with epileptic seizures.
a. It was reported that " $20 \%$ of families with epileptic children said their pets have the ability to anticipate a seizure. . ." Was this information most likely obtained with an open or a closed question? Is the variable of interest quantitative or categorical?
b. It was also reported that "anticipation time ranged from 10 seconds to five hours, with an average of 2.5 minutes." Was this information most likely obtained with an open or a closed question? Is the variable of interest quantitative or categorical?
c. Furthermore, "responses from the 122 families in the survey reveal protective behaviors by a variety of dogs. For example: A sheltie-spitz mixed breed sits on a child before a drop attack, a seizure that causes an epileptic to fall to the ground. An Akita keeps a young girl away from the stairs 15 minutes before a seizure. And a Great Pyrenees attaches itself to a 3 -year-old, forgoing even food and water, hours before the girl has a convulsion." What is the name for this type of evidence?

"Don't Count Out Prostitutes" reports that "surveys for 15 years have shown that men have more sex partners than women, an illogical conclusion that has puzzled experts. The General Social Surveys, conducted by the University of Chicago, and the National Health and Social Life Survey, found that men were claiming up to $74 \%$ more partners than women. ${ }^{33}$
a. One explanation for the discrepancy would be that there is bias resulting from which one of these difficulties: complicated questions, sensitive questions, or hard-to-define concepts?
b. Is the variable of interest quantitative or categorical?

Tell whether the following questions are open or closed.
a. Patti LaBelle sang, "Voulez vous coucher avec moi ce soir?" (Do you want to sleep with me tonight?)
b. Beck sang, "Soy un perdedor. I'm a loser baby so why don't you kill me?"

Tell whether the following pairs of questions are open or closed.
a. Bill Withers sang, "Who is he (and what is he to you)?"
b. The Beatles sang, "Will you still need me? Will you still feed me when I'm sixty-four?"

Find song lyrics that pose a question, and tell whether it is open or closed.

Tall Tale? A Study Suggests That Spring Babies Have More Stature" reports that spring babies were found to be taller, by 0.23 of an inch, than those born in the fall. Tell why this had to be an observational study, not an experiment.

A Swedish researcher proposed a theory that links the production of shoes to the prevalence of schizophrenia: "Heeled footwear began to be used more than 1,000 years ago, and led to the occurrence of the first cases of schizophrenia . . .

An article entitled "Breast Milk Benefit" states: "In a review of 212 premature births from 1992 and 1993, the researchers found that $29 \%$ of the infants fed human milk acquired infections versus $47 \%$ of the babies fed formula. "35
a. Was this an observational study or an experiment?
b. Which of these would be most worrisome here: confounding variables, people's behaviors affected by their awareness of participating in a prospective study, or people's faulty memories in a retrospective study?

An article entitled "Breast-Feeding Benefits Bolstered" explains that "hospitals were assigned at random to institute a breastfeeding program . . . in which doctors and midwives gave instruction and counseling. The other hospitals . . . provided the usual obstetric care. ${ }^{36}$ Results showed that "about $9 \%$ of the infants who had been in the breast-feeding program had at least one intestinal infection in the first year, compared with about $13 \%$ in the control group."
a. Was this a paired or a two-sample study?
b. Was this an observational study or an experiment?
c. Are the results from this study more convincing than those from the study described in the previous exercise because of a larger sample size, because of a bigger difference in percentages acquiring infections for breast-fed versus bottle-fed children, or because of a better study design?

"Boozy Bees May Offer Clues About Pickled People" describes a study in which bees were given various amounts of alcohol, while

A Quaker Oatmeal box reported the following: "100 people in Lafayette, Colorado volunteered to eat a good-sized bowl of oatmeal for 30 days to see if simple lifestyle changes-like eating oatmealcould help reduce cholesterol. After 30 days, 98 lowered their cholesterol. With these great results, the people in Lafayette proved to themselves that simple changes can make a real difference. ${ }^{37}$
a. Was this an observational study or an experiment?
b. There are two variables of interest. What was the explanatory variable?
c. Was the explanatory variable categorical or quantitative?
d. What was the response variable?
e. Was the response variable categorical or quantitative?
f. Describe the treatment group.
g. Was there a randomized assignment to treatment or control? Explain.
h. Describe other possible reasons, besides eating oatmeal, that could account for the participants' cholesterol being lowered.
i. Which one of these is more problematic here: experimenter effect or Hawthorne effect?
j. Would Quaker Oats' results have been more convincing if it had used 1,000 instead of 100 volunteers?
k. Describe details of a study design that might be more effective than this one in establishing whether or not oatmeal lowers cholesterol.

Suppose that the effect of consuming oatmeal on cholesterol level is thought to depend to some extent on two additional variables-gender and degree of fitness. For which of these two variables would it be easier to separate the data when exploring the relationship between oatmeal consumption and cholesterol level?

In which study would noncompliance be more of a problem: a study that asks participants to eat oatmeal for 30 days, or a study that asks participants to use pesticides for 2 years?

In which study could the experimenter effect be more of a problem: a study in which the response of interest is cholesterol levels, or a study in which the response of interest is how sociable a bee's behavior is?

In which study could the experimenter effect be more of a problem: a study in which the response of interest is people's feelings of self-esteem, or a study in which the response of interest is the size of a wound, in square centimeters?

Which of these studies would cost more money to carry out: the study that tests if eating oatmeal every day for a month lowers cholesterol levels, or the study that tests children's absorption of pesticides by paying their parents to use certain pesticides over the course of 2 years?

Which of these studies would cost more money to carry out: the study that tests whether writing has an effect on healing when tissue has been removed from the upper arm under anesthesia, or the study that tests whether people's self-esteem is lowered if they are asked to think about times when they had bad hair?

A study at Northwestern University "evaluated 67 normal, mentally healthy, suburban males initially in 1962 , when the boys were 14, and in 1997 , when the men were $48 .{ }^{38}$ One interesting finding was that " $44 \%$ of the men 'remembered' that when they were teenagers, they believed that it was OK to have sex during high school. However, only $15 \%$ of the participants felt this way when they were questioned as teenagers."

Find (and hand in) an article or Internet report about a sample survey. Tell if each variable of interest is quantitative or categorical. Tell what is the suggested population of interest. Then tell how the individuals were selected and whether or not you believe they adequately represent the population of interest. Discuss whether there are any clear sources of bias. Were the questions open or closed?

Find (and hand in) an article or report about an observational study. Tell what the variables of interest are, whether they are quantitative or categorical, and which is explanatory and response (if there are two variables). Are there any potential confounding variables that should have been controlled for?

Find (and hand in) an article or report about an experiment. Tell what the variables of interest are, whether they are quantitative or categorical, and which is explanatory and response. Describe the subjects, treatments, whether or not the subjects were blind, whether or not the researchers were blind, and whether there are any obvious problems with the experiment's design.

Use results of Exercises 3.16, 3.17, 3.18 and 3.21 and relevant findings from the Internet to make a report on gun control that relies on statistical information.