1. and 2. Epidemiologists in Britain noted that certain cancers were more frequent among residents living near nuclear power plants than in the population at large.
1. Indicate which of the following models are consistent with this correlation? (You do not need to worry about the subsequent epidemiological studies discussed in class to answer this question):
2. Now suppose that we had randomly assigned where people live in Britain over the last 100 years, and that we still observed that residents living near nuclear power plants had higher-than-average cancer rates. Which of the following models would be consistent with this correlation?
3. Epidemiologists in the US have discovered that certain cancers are found in residents living near gasoline tank farms. What control group is most appropriate and useful for evaluating the hypothesis that gasoline tank farms cause cancer?
4. Mark all of the following statements about controls that are correct.
5. Mark the letters that describe design features indicated in the following study description. Base your answer only on the information given. That is, only mark a design feature if it is explicitly present in the problem description. If none apply, do not mark any.
Experimental Design Features
5. As a source of temporary employment, you go into sales. You want to find out whether you are a more effective salesperson wearing a blue than a gray suit. You thus purchase one suit of each color. To avoid potential confounding factors, you decide to wear the blue suit on Monday, Wednesday, and Friday (the gray one on Tuesday & Thursday) if the week begins on an odd-numbered day of the month, whereas you reverse the order if the week begins on an even-numbered day. At the end of two months, you compare the number of sales you made wearing the blue versus gray suit.
(D) is optional; that is, it would be marked correct regardless of whether you marked it. The customers do not know about the experiment, even though you do.
6. to 8. For these questions, mark all correct summaries of the information in the question. Base your answer only on the information given in the question. The possible answers are:
6. A scientist investigates the effect of maternal smoking during pregnancy on the fetus by determining the correlation between the infant's weight at birth, and the average number of cigarettes the mother smoked per day during pregnancy. The study involves twenty haphazardly chosen mothers and their infants, and includes non-smoking mothers as well as mothers who smoke. All the participants in the study were patients at a public clinic that specialized in caring for young, single mothers.
7. A college professor wants to determine if her grading of essay exams is biased. To determine this, she has the T.A. randomly divide the fifty midterm exams into two groups. For exams in the first group, the T.A. removes the students' name and replaces it with a code (such as 33-2A). For exams in the second group, the T.A. leaves the students' names on the exam. She then grades the exams, and compares the scores in the two groups.
8. The Austin police chief suspects that juvenile delinquents frequently live in households with only one parent. She asks her assistant to obtain some further information about this. He obtains the records of all one hundred children over the age of 15 currently at the juvenile detention center (i.e. jail), and he discovers that 30% of them come from a household with only one parent.
9. You live several blocks from a large refinery. One day, a neighbor gives you some data showing that, from a survey of all 2,400 people living within 1/2 mile of the plant, 13 have developed leukemia during the last year. On the basis of these data, your neighbor claims that the refinery is a health hazard in the community and asks your support in petitioning to have it closed. Which single additional feature is needed the most before you can decide if the refinery may be a health hazard?
10. The Monty Python tape shown in lecture investigated the relative intelligence of humans and penguins. Which group of humans considered in the tape was the best control for this comparison:
11. and 12. In 1995 the Texas Boll Weevil Eradication Foundation initiated a program to spray cotton crops in south Texas with malathion. Because of this malathion spraying program, the cotton crops in south Texas received a much heavier dose of malathion in 1995 than they did in previous years. Unfortunately, the 1995 south Texas cotton crop was only 1/6 the size of the crop in previous years. Additionally scientists found that the 1995 cotton crop suffered from heavy infestations of the beet armyworm.
It was hypothesized that the large reduction in the size of the 1995 cotton crop was caused by the large beet armyworm infestation, and that this infestation occurred because malathion killed the wasps and spiders that would have otherwise preyed on beet armyworms. In 1995, the Texas Boll Weevil Eradication Foundation sprayed cotton crops with malathion only in the United States. Right across the Rio Grande river there are some Mexican cotton farmers who did not spray with malathion. The yield of these Mexican cotton crops was as high in 1995 as in past years, and these crops were not heavily infested with beet armyworm. In any given year, these Mexican crops are subjected to nearly the same temperature and rainfall as are the south Texas crops, however different years have different temperature and rainfall. Late season vegetables are grown in south Texas, but not in Mexico; their presence or absence may affect the size of the cotton crop.
Available Data: Crop Sizes in Different Years and Countries
In answering the following questions, rely on the information presented in the preceding paragraphs. For each question, you may mark 0, 2, or more than 2 answers. (Thus, under no circumstances should you mark a single answer.) If more than one comparison is informative, choose the most informative comparison among the choices given; however, if more than one data set is required, then you should mark all data sets that are required. Any comparison you mark must be obviously relevant to the model being investigated; if no comparisons are relevant, mark none. In answering each question, consider the effects of late season vegetables, weather, and malathion on cotton production, but do not consider any other factors.
11. Which crop yields should you compare to investigate the model that U.S. cotton crops are smaller than Mexican crops because late season vegetables are grown in the U.S. but not in Mexico.
12. Which crop yields should you compare to investigate the model that the weather in 1995 was particularly unfavorable for cotton, as compared to 1994.