Chapter 7 Solutions

7.1 This an observational experiment. The subjects were not assigned to their usage of cell phones. The explanatory variable is cell phone usage and the response variable is incidence of brain cancer.
7.3 Adolescents whose parents are rarely home may watch more television. Adolescents who are not involved in as many social activities may way watch more television. Adolescents whose parents watch more television may watch more television as well.
7.4 The population is not well defined in the question. If it is a national opinion poll, then it was probably done via telephone, so the population would be all adults who have a residential phone and are home at during the time that the survey was being conducted.
7.6 The population is all regulators in the last shipment.
7.7: a) The sample size for this poll is 14793 + 1778 + 2566 + 988 = 20125.; b) There was no systematic effort to sample. This sample was a voluntary response sample.
7.9 Two digit numbers were assigned to the names starting with the number 01 in the upper left and proceeding down each column (taking columns from left to right). Using the random digits from Table B starting at line 139, the six managers to be interviewed are (in order of selection), Bonds, Fleming, Liao, Naber, Goel, and Gomez.
7.11Starting at line 101 of Table B using two digit numbers since there are more than 10 students, the first 4 two-digit numbers less than or equal to 30 are, 19, 22, 05, and 13. Starting at line 102 and using the first 2 one-digit numbers since there are 10 teachers gives 7 and 3. With students and teachers assigned numbers starting with 01 (0) and moving from left to right and then down, results in students O'Brien, Hernandez, Reinmann, and Jones and faculty Gupta and Kim being selected.
7.13 The period with the higher percentage of "ring-no-answer" results was probably between July 1 and August 31. People are more likely to be on vacation during that time. High rates of nonresponse make surveys less reliable because there may be a segment of the population that is more likely to be nonreponsive and also similar in other important ways.
7.15 The firm asks more people in order to get more precise estimates of what the population is actually thinking about the presidential election.
7.16 This is an observational study because the subjects were not assigned their level of alcohol consumption. The explanatory variable is alcohol consumption and the response variable is survival rate 4 years after a heart attack.
7.18: a) The study was observational because researchers used the information of previous records without manipulating the explanatory variable directly with some design.; b) Some other variables that may confound the results of this investigation include the type of surgery, the purpose of the surgery, drug interactions (if the patient is taking other drugs), and patient sensitivity to anesthetic (allergic reactions).
7.21 The population is small businesses. The sample is the subset of the 150 eating-and-drinking establishments selected from the yellow-pages that choose to respond. Note that this sample is not random (nor does it seem likely to be representative) given that the possible businesses that could be selected did not include the entire population.
7.24 When people know they are supposed to do something, either because it is against the law not to, or because of other social pressures, some will tend to say that they do that thing, if asked. So when people were asked in this survey whether they always wore their seatbelt, there was a tendency for some to say that they did even though they did not. There will probably be bias in the same direction in most surveys about seatbelt use.
7.25 The population is black residents of Miami. The sample is the adult residents of the 300 addresses visited who choose to respond. The survey results are likely to be biased because the questions are about the perception of police officers and a police officer (in uniform) is asking the questions. People will be less likely to respond negatively as a result.
7.30 (45,956-5029)/45,956 = 89% nonresponse rate.
7.32 Each student has a 10% chance of being interviewed. The sample is not a SRS because not every sample has the same probability of being selected. For example, there is a 0% chance that all students are over age 21 using this method. This sample uses a stratified random sampling approach.
7.33: a) This is not true. On average there will be four 0's in each row of 40 digits, but it is often the case that there will not be exactly 4 in a row.; b) It is true that each pair of digits has a 1/100 chance of being 00.; c) The digits 0000 can appear as a group. Random numbers have no memory of what came before, so it is possible for a digit to be repeated many times (though it is unlikely; in this case, only a 1/10000 chance).
7.40: a) Should the city of Sitka ruin the characteristics that make it so attractive to residents and visitors alike by building a deep water dock near downtown?; b) Should the State of Alaska not discontinue the current method for calculating dividends by changing the formula to five percent of the total value, rather than using only the surplus (after inflation proofing)?