Exam Prep Causation: Can We Say What Caused The Effect? Ch.4

Chapter 4

Introduction to Statistical Investigations Test Bank

Note: TE = Text entry TE-N = Text entry - Numeric

Ma = Matching MS = Multiple select

MC = Multiple choice TF = True-False

DD = Drop-down

CHAPTER 4 LEARNING OBJECTIVES

4-1: Identify confounding variables in observational studies.

4-2: Explain how studies using random assignment are able to draw cause-and-effect conclusions.

Section 4.1: Association and Confounding

4.1-1: Calculate and interpret conditional proportions.

4.1-2: Interpret conditional proportions as to whether they give any indication of an association between the explanatory and response variables.

4.1-3: Identify which variable is the explanatory variable and which is response in a study involving two variables.

4.1-4: Identify potential confounding variables and explain how they provide an alternative explanation for the observed association between the explanatory variable and the response variable.

4.1-5: Draw a diagram to show how the confounding variable provides an alternative explanation for the observed association between the explanatory variable and the response variable.

Questions 1 and 2: According to the 2020 Bureau of Transportation Statistics^[1], 55.7% of flights arrive on-time. Of those flying Delta, 67.6% of flights arrive on time. However, only 47.0% of flights arrive on time for those flying Southwest.

1. Which of the following values are conditional probabilities in this scenario? Select all that apply.
   1. 55.7%
   2. 67.6%
   3. 47.0%
   4. 2020
2. Does there appear to be an association between the airline company (Delta or Southwest) and if a flight arrives on time?
   1. No, since 55.7% of all flights arrive on time.
   2. Yes, since the percent of flights that arrive on time differs between airlines.
   3. Yes, since the percent of flights that arrive on time for each airline is not equal to 50%.
   4. Yes, since there are other airlines besides Delta and Southwest.
3. January 22, 2016, NPR news headline reads “Why Poverty May Be More Relevant Than Race For Childhood Obesity.” The article states, “At first glance it looked like childhood obesity was more common among African Americans or Hispanics... When they accounted for poverty, though, the trend vanished.” The study collected data on a random sample of K-12 students in Massachusetts. In this study, whether or not a child lives in poverty is an example of a
   1. quantitative variable.
   2. response variable.
   3. confounding variable.
   4. normal variable.

Questions 4 through 6: A study published in 2007 by Christopher Johnson, professor of music education and music therapy at the University of Kansas, revealed that students in elementary schools who took private music lessons scored around 20 percent higher in math scores on standardized tests, compared to students who did not take private music lessons.

1. What is the explanatory variable?
   1. Music therapy
   2. Whether a student took private music lessons
   3. Math score on standardized test
   4. Grade level
2. What is the response variable?
   1. Music therapy
   2. Whether a student took private music lessons
   3. Math score on standardized test
   4. Grade level
3. Which of the following could be a potential confounding variable?
   1. Household income
   2. Socioeconomic status
   3. Whether the parents hired a math tutor
   4. All of the above

Questions 7 and 8: Babies born with low birth weights (less than 2500 grams) are at an increased risk for many infant diseases. Researchers in North Carolina collected data to see what variables may influence the birth weight (in grams) of a child, including whether the mother drank alcohol during pregnancy, whether the mother smoked during pregnancy, the mother’s age (years), the gestation of the pregnancy (number of weeks from conception until birth), the mother’s race, the length of the birth (hours), and several others. The following plot was created to summarize some of the data collected.

1. What is the response variable in this study?
   1. Whether the mother smoked during pregnancy
   2. Gestation of the pregnancy
   3. The mother’s age
   4. Birth weight
2. Explain why whether the mother drank alcohol during pregnancy is a confounding variable when trying to determine if smoking increases the risk of low birth weight.
   1. Drinking alcohol may be associated with low birth weight, and mothers who drink may be more likely to also smoke.
   2. Drinking alcohol may be associated with high birth weight, and mothers who drink may be more likely to also smoke.
   3. Drinking alcohol may be associated with low birth weight, and mothers who drink may be less likely to also smoke.
   4. None of the above.

Questions 9 through 11: The following table shows the responses from a sample of 680 people in the General Social Survey to the question, “Do you sometimes drink more than you think you should?”

1. What proportion of males responded that they drink more than they should?
   1. 151/243
   2. 151/328
   3. 243/680
   4. 151/680
2. What proportion of those that responded they drink more than they should are female?
   1. 151/243
   2. 92/243
   3. 92/680
   4. 352/680
3. Does there appear to be an association between gender and whether a person responds that they drink more than they should?
   1. Yes, since there are a different number of people of each gender.
   2. Yes, since the percent that responded that they drink more than they should differs from 50%.
   3. Yes, since the proportion who responded they drink more than they should differs between males and females.
   4. No, there does not appear to be an association between these two variables.

Questions 12 and 13: A study was done to compare the lung capacity of coal miners to the lung capacity of farm workers. The researcher studied 200 workers of each type. Other factors that might affect lung capacity are smoking habits and exercise habits. The smoking habits of the two worker types are similar, but the coal miners generally exercise less than the farm workers.

1. Which of the following is the explanatory variable in this study?
   1. Exercise
   2. Lung capacity
   3. Smoking or not
   4. Occupation
2. Which of the following is a confounding variable in this study?
   1. Exercise
   2. Lung capacity
   3. Smoking or not
   4. Occupation

Questions 14 and 15: In a study on sleep habits, it was found that college students who got at least 7 hours of sleep the night before a test performed better on exams than college students who got less than 7 hours of sleep.

1. Identify the explanatory and response variables in this study.
   1. The explanatory variable is exam score, and the response variable is whether a student got at least 7 hours of sleep.
   2. The explanatory variable is whether a student got at least 7 hours of sleep, and the response variable is exam score.
   3. The explanatory variable is sleep habits, and the response variable is performing better on exams.
   4. The explanatory variable is performing better on exams, and the response variable is sleep habits.
2. In the following diagram, fill in the blanks to describe a potential confounding variable in this study.

Drop-down menu for (1) and (2):

• Enrolled in more credit hours
• Enrolled in less credit hours
• Perform better on exams
• Perform worse on exams

Section 4.2: Observational Studies Versus Experiments

4.2-1: Identify a study as observational or experimental.

4.2-2: Explain that random assignment gives us the ability to draw cause–effect conclusions because it ensures that treatment groups have similar characteristics.

4.2-3: Identify whether a study uses random assignment and/or random sampling and the implications of these design decisions on the conclusions that can be drawn.

4.2-4: Describe what a block study design is and the benefits of using a blocking variable.

1. January 22, 2016, NPR news headline reads “Why Poverty May Be More Relevant Than Race For Childhood Obesity.” The article states, “At first glance it looked like childhood obesity was more common among African Americans or Hispanics... When they accounted for poverty, though, the trend vanished.” The study collected data on a random sample of K-12 students in Massachusetts. Does this study design allow us to generalize results to the population of all Massachusetts K-12 students?
2. Yes, since it is an observational study.
3. Yes, since it is a randomized experiment.
4. Yes, since the study used a random sample from the population.
5. No.
6. A study published in 2007 by Christopher Johnson, professor of music education and music therapy at the University of Kansas, revealed that students in elementary schools who took private music lessons scored around 20 percent higher in math scores on standardized tests, compared to students who did not take private music lessons. Is this a randomized experiment or an observational study?
   1. Randomized experiment, since researchers gave students standardized tests.
   2. Observational study, since whether the student took private music lessons was not randomly assigned.
   3. Observational study, since math scores were not randomly assigned.
   4. Observational study, since this was not a random sample.
7. Babies born with low birth weights (less than 2500 grams) are at an increased risk for many infant diseases. Researchers in North Carolina collected data to see what variables may influence the birth weight (in grams) of a child, including whether the mother drank alcohol during pregnancy, whether the mother smoked during pregnancy, the mother’s age (years), the gestation of the pregnancy (number of weeks from conception until birth), the mother’s race, the length of the birth (hours), and several others. Is this a randomized experiment or an observational study?
   1. Randomized experiment, since researchers measured more than two variables.
   2. Observational study, since birth weight cannot be randomly assigned.
   3. Observational study, since whether the mother drank alcohol during pregnancy was not randomly assigned.
   4. Observational study, since this was not a random sample.

Questions 19 and 20: A randomized experiment was done by randomly assigning each participant either to walk for half an hour three times a week or to sit quietly reading a book for half an hour three times a week. At the end of a year the change in participants' blood pressure over the year was measured, and the change was compared for the two groups.

1. This is a randomized experiment rather than an observational study because:
   1. Blood pressure was measured at the beginning and end of the study.
   2. The two groups were compared at the end of the study.
   3. The participants were randomly assigned to either walk or read, rather than choosing their own activity.
   4. A random sample of participants was used.
2. If a statistically significant difference in blood pressure change at the end of a year for the two activities was found, then:
3. It cannot be concluded that the difference in activity caused a difference in the change in blood pressure because in the course of a year there are lots of possible confounding variables.
4. Whether or not the difference was caused by the difference in activity depends on what else the participants did during the year.
5. It cannot be concluded that the difference in activity caused a difference in the change in blood pressure because it might be the opposite, that people with high blood pressure were more likely to read a book than to walk.
6. It can be concluded that the difference in activity caused a difference in the change in blood pressure because of the way the study was done.
7. Researchers randomly assigned patients with severe acne to one of two treatments: the standard topical antibiotic (our control) or a new antibiotic taken orally (the treatment). Six months later an evaluation of skin problems was made by a doctor who did not know which medication subjects had been given. What is the primary advantage of random assignment of the treatment?
8. It ensures that the study participants are representative of the larger population.
9. It allows inference back to a population of patients with acne.
10. It allows the researchers to infer that any difference in outcome was due to the treatments.
11. It ensures that the subject doesn't know which treatment they are getting.
12. Recently, a Montana newspaper asked people to go to their website and respond yes or no to the question: “Should federal protections on grizzly bears be lifted?" Of 933 opinions, 58.1% said “Yes", and 41.9% said either “No" or “Uncertain". Is it valid to conclude from the poll that over half of all Montana residents favor lifting the protections?
13. Yes, because 58.1% is far enough above 50% to conclude that the true proportion is more than one-half.
14. Yes, because the sample size is large enough to represent all 75,000 Montana residents.
15. No, because the sample is not likely to be representative of the population.
16. No, because the sample is too small given the population of 75,000 Montana residents.
17. A Pew Research Center poll randomly selected and surveyed 2,002 US adults in the 25-32 age group and asked about their annual income. They found strong evidence that the unemployment rate for people without a college degree was higher than the unemployment rate of those with a college degree. Can the researchers conclude that college is the cause of the decrease in unemployment among 25-32 year old U.S. adults?
18. No, this study does not allow causation to be inferred.
19. Yes, the statistically significant result allows causation to be inferred.
20. Yes, there is strong evidence that college graduates have a lower unemployment rate.
21. No, the sample size is too small to allow causation to be inferred.

Questions 24 and 25: Sickle-cell disease is a painful disorder of the red blood cells. To investigate whether drug A can reduce the pain associated with sickle-cell disease, a study by the National Institutes of Health randomly assigned 150 sickle-cell sufferers to receive the drug. Placebos were given to another 150 sickle-cell sufferers. Great care was used to ensure that the 300 participants did not know if their pill contained the drug. At the end of the treatment period, the researchers counted the number of episodes of pain reported by each subject.

1. What type of study is this?
2. Observational study
3. Randomized experiment
4. Which of the following principles was NOT used in this study?
5. Control group
6. Blinding
7. Randomization
8. Blocking
9. True or false: A matched pairs design will have less variability in the differences in mean response between two groups than a completely randomized design.
10. A study published in 2007 by Christopher Johnson, Professor of Music Education and Music Therapy at the University of Kansas, revealed that students in elementary schools with superior music education programs scored around 20 percent higher in math scores on standardized tests, compared to schools with low-quality music programs. Is this an observational study or a randomized experiment?
    1. Observational study
    2. Randomized experiment

Questions 28 through 30: In a study on treatments for anorexia, researchers recruited 30 teenage girls with anorexia. The girls were ranked by weight prior to the study. Of the heaviest three girls, one was randomly assigned to a “family” therapy, one to a “cognitive/behavioral” therapy, and one to a control group. The same was done for the next three heaviest girls, down to the three lightest girls. After five weeks of treatment, researchers weighed the participants. Researchers would like to determine which, if any, of the treatments are most effective in treating anorexia.

1. What type of study design did the researchers use?
   1. Completely randomized design
   2. Matched pairs design
   3. Blocking study design
   4. Observational study
2. If the researchers found that there were statistically significant differences in average weight across the three treatments, would they be justified in concluding that the treatments caused the changes in weights?
   1. No, since we do not know which treatment was most effective.
   2. No, since weight was not randomly assigned.
   3. Yes, since therapy group was randomly assigned.
   4. Yes, since this was an observational study.
3. Can the results from this study be generalized to the population of all teenage girls with anorexia?
   1. No, since the participants in the study were volunteers.
   2. No, since it is an observational study.
   3. Yes, since it is a randomized experiment.
   4. Yes, since the treatments were randomly assigned.

1. https://www.transportation.gov/sites/dot.gov/files/2020-06/June%202020%20ACTR.pdf ↑