,Solution manual for Mind on Statistics , 6th Edition
Jessica M. Utts
Notes
1- The file is chapter after chapter.
2- We have shown you 10 or five pages.
3- The file contains all Appendix and Excel
sheet if it exists.
4- We have all what you need, we make
update at every time. There are many
new editions waiting you.
5- If you think you purchased the wrong file
You can contact us at every time, we can
replace it with true one.
Our email:
,Mind on Statistics
6th Edition
Complete Solutions Manual
© 2022 Cengage Learning, Inc.
, © 2022 Cengage Learning, Inc. ISBN: 978-133779484-8
Cengage
Unless otherwise noted, all content is © Cengage. 200 Pier 4 Boulevard
Boston, MA 02210
ALL RIGHTS RESERVED. No part of this work covered by the USA
copyright herein may be reproduced or distributed in any form
or by any means, except as permitted by U.S. copyright law, Cengage is a leading provider of customized learning
without the prior written permission of the copyright holder. solutions with employees residing in nearly 40 different
countries and sales in more than 125 countries around
the world. Find your local representative at:
www.cengage.com.
For product information and technology assistance, contact us at
Cengage Customer & Sales Support, To learn more about Cengage platforms and services,
1-800-354-9706 or support.cengage.com. register or access your online learning solution, or
purchase materials for your course, visit
For permission to use material from this text or product, submit www.cengage.com.
all requests online at www.cengage.com/permissions.
NOTE: UNDER NO CIRCUMSTANCES MAY THIS MATERIAL OR ANY PORTION THEREOF BE SOLD, LICENSED, AUCTIONED,
OR OTHERWISE REDISTRIBUTED EXCEPT AS MAY BE PERMITTED BY THE LICENSE TERMS HEREIN.
READ IMPORTANT LICENSE INFORMATION
Dear Professor or Other Supplement Recipient: electronically must be through a password-protected site, with all
copy and download functionality disabled, and accessible solely by
Cengage Learning has provided you with this product (the your students who have purchased the associated textbook for the
“Supplement”) for your review and, to the extent that you adopt Course. You may not sell, license, auction, or otherwise
the associated textbook for use in connection with your course redistribute the Supplement in any form. We ask that you take
(the “Course”), you and your students who purchase the textbook reasonable steps to protect the Supplement from unauthorized
may use the Supplement as described below. Cengage Learning use, reproduction, or distribution. Your use of the Supplement
has established these use limitations in response to concerns indicates your acceptance of the conditions set forth in this
raised by authors, professors, and other users regarding the Agreement. If you do not accept these conditions, you must return
pedagogical problems stemming from unlimited distribution of the Supplement unused within 30 days of receipt.
Supplements.
All rights (including without limitation, copyrights, patents, and
Cengage Learning hereby grants you a nontransferable license to trade secrets) in the Supplement are and will remain the sole and
use the Supplement in connection with the Course, subject to the exclusive property of Cengage Learning and/or its licensors. The
following conditions. The Supplement is for your personal, Supplement is furnished by Cengage Learning on an “as is” basis
noncommercial use only and may not be reproduced, or without any warranties, express or implied. This Agreement will
distributed, except that portions of the Supplement may be be governed by and construed pursuant to the laws of the State
provided to your students in connection with your instruction of the of New York, without regard to such State’s conflict of law rules.
Course, so long as such students are advised that they may not
copy or distribute any portion of the Supplement to any third party. Thank you for your assistance in helping to safeguard the
Test banks and other testing materials may be made available in integrity of the content contained in this Supplement. We trust you
the classroom and collected at the end of each class session, or find the Supplement a useful teaching tool.
posted electronically as described herein. Any material posted
Printed in the United States of America
Print Number: 01 Print Year: 2022
, CHAPTER 1
EXERCISE SOLUTIONS
1.1 a. The fastest speed was 150 miles per hour.
b. The slowest speed driven by a male was 55 miles per hour.
c. 1/4 of the females reported having driven at 95 miles per hour or faster. Notice that
95 mph is the upper quartile for females. By definition, about 1/4 of the values in a dataset are greater
than the upper quartile.
d. 1/2 of the females reported having driven 89 mph or faster. Notice that 89 mph is the median value.
e. 1/2 of 102 = 51 females have driven 89 mph or faster.
Note: For parts (d) and (e) the answer would have to be adjusted if there were any females who reported 89
as their value, but from the data on page 4 we can see that there were not. Because there were no “ties” with
the median, we know that exactly half of the values fall above it and the other half fall below it.
1.2 a. The median height is 65 inches.
b. Range = tallest – shortest = 71 – 59 = 12 inches.
c. The interval from 59 to 63.5 inches contains the shortest 1/4 of these students. This interval is from the
minimum to the lower quartile.
d. The interval from 63.5 to 67.5 inches contains the middle 1/2 of these students. This is an interval from
the lower quartile to the upper quartile.
1.3 a. The observed rate of cervical cancer in Vietnamese American women is 86 per 200,000. This could also
be expressed as 43 per 100,000 or 4.3 per 10,000, and so on. In decimal form, it is .00043.
b. The risk of developing cervical cancer for Vietnamese American women in the next year is 86/200,000
= .00043.
c. The rate of 86 per 200,000 is based on past data and tells us the number of Vietnamese American
women who developed cervical cancer out of a population of 200,000. The risk utilizes the rate from the
past to tell us the future likelihood of cervical cancer in other Vietnamese American women.
1.4 a. The base rate is about 13 in 1000, or about .013.
b. The risk for men who smoke is just over 13 times the rate for non-smokers, or about .169.
1.5 a. All teens in the United States at the time the poll was taken.
b. All teens in the United States who had dated at the time the poll was taken.
1.6 A population is a collection of all individuals of interest while a sample is a subset of the population of
interest, for which measurements are taken in a study. In Case Study 1.6, the population of interest is
probably all men, and possibly women as well. However, the sample consisted of 22,071 male physicians
who volunteered for the study, so the population to which the results apply is all men similar to them.
1.7 a. All adults in the United States at the time the poll was taken.
1
b. = .031 or 3.1%.
1048
c. 34% ± 3.1%, or 30.9% to 37.1%.
1.8 a. The population is probably all Canadians who were eligible to participate in the survey (which is
probably all adults with telephones).
b. There were 2000 people in the sample.
1 1
c. = = .022 or 2.2%.
n 2000
d. In the sample, 16% viewed immigration as having a negative impact. The interval 16% ± 2.2%, or
13.8% to 18.2%, is 95% certain to cover the true percent of Canadians who viewed immigration as
having a negative impact (at the time of the poll).
Page 1 of 6
,Chapter 1 Solutions, 6th ed. Page 2 of 6
1 1
1.9 Solve for n in the equation = .05 = . Answer is n = 400 teenagers.
n 20
1
1.10 Solve for n in the equation = .10. Answer is n = 100.
n
1.11 a. This is an example of a self-selected or volunteer sample. Magazine readers voluntarily responded to the
survey and were not randomly selected.
b. These results may not represent the opinions of all readers of the magazine. The people who respond
probably do so because they feel stronger about the issues (for example, violence on television or
physical discipline) than the readers who do not respond. So, they may be likely to have a generally
different point of view than those who do not respond.
1.12 The exercise did not specify what the survey is about, but no matter what, the survey is based on a
self-selected sample, and people who feel strongly about the issues and/or who have extra time are more
likely to respond. The results will not be representative of all students who use the cafeteria.
1.13 a. Randomized experiment (because students were randomly assigned to the two methods).
b. Observational study (because people cannot be randomly assigned to smoke or not).
c. Observational study (because people cannot be randomly assigned to be a CEO or not).
1.14 a. Randomized experiment, because students were randomly assigned to receive Vitamin C or placebo.
b. Observational study, because the patients were not randomly assigned to do anything. (Note that a
random sample is not the same thing as random assignment.)
c. Randomized experiment, because participants were randomly assigned to meditation or low-fat diet.
1.15 Answers will vary, but one possibility is general level of activity. It is likely to differ for elderly people
who attend church regularly and those who don’t, and it is also likely to affect blood pressure. So it might
partially explain the results of this study.
1.16 a. Number of courses might be a confounding variable. Students taking many courses may sleep less due
to the amount of work involved and may not do as well in school due to the load.
b. Weight is not likely to be a confounding variable. Weight is probably not related either to amount of
sleep or to grades.
c. Hours spent partying might be a confounding variable. Students who party a lot may sleep less, and may
also get lower grades because they’re not studying.
1.17 You would need to know how large the difference in weight loss was for the two groups. If the difference
in weight loss is very small (but not 0), it could be statistically significant, but not have much practical
importance.
1.18 Statistical significance is when there is a relationship or difference that is large enough to be unlikely to
have occurred in the sample if there was no relationship or difference in the population of interest. Practical
significance occurs when the relationship or difference is large enough to be important or meaningful in a
“real world” sense. A result can be statistically significant, but not practically significant. This is especially
likely to occur in studies with very large sample sizes.
1.19 You would want to know how many different relationships were examined. If this result was the only one
that was statistically significant out of many examined, it could easily be a false positive.
1.20 A false positive occurs when a relationship or difference is said to be statistically significant based on
examining information from a sample, but in fact there is no relationship or difference in the population.
1.21 The placebo group estimates the baseline rate of heart attacks for men not taking aspirin. So, the estimated
baseline rate of heart attacks is 189/11,034, which is about 17 heart attacks per 1000 men or 17/1000. (See
Table 1.1 for the data.)
1.22 a. The amount of exercise per week is similar for men and women except that there are a few high values
for the men. The dotplot follows.
,Chapter 1 Solutions, 6th ed. Page 3 of 6
Figure for Exercise 1.22a
Gender
m
w
0 100 200 300 400 500 600
Exercise per week (minutes)
b. Women, median = 190 min. Men, median = 180 min.
To find the median, put the data in order first.
For women, the ordered list of data is:
0, 0, 0, 60, 60, 70, 100, 180, 200, 240, 240, 270, 300, 360, 360, 450
The number of women is even (16), so the median is the average of the middle two values in the ordered
list. These middle two values, underlined and bold in the above list, are 180 and 200 and their average is
190.
For men, the ordered list of data is:
0, 0, 14, 60, 90, 120, 180, 240, 300, 300, 360, 480, 600
The number of men is odd (13), so the median is the middle value in the ordered data. This value,
underlined and bold in the list above, is 180.
c. Although the median response is different for women and men, the difference is only 10 minutes. The
weekly amount of exercise is about the same for the samples of women and men.
1.23 a.
Minutes of Exercise per Week
Median 180
Quartiles 37 330
Extremes 0 600
To determine the summary, first write the responses in order from smallest to largest.
The ordered list of data is:
0, 0, 14, 60, 90, 120, 180, 240, 300, 300, 360, 480, 600
Minimum = 0 min.
Maximum = 600 min.
Median = 180 min. (middle value in the ordered list)
Lower quartile = 42 min. It is the median of the values smaller than the median.
These are 0, 0, 14, 60, 90, 120.
Median of these six values is (14 + 70)/2 = 42.
Upper quartile = 330 min. It is the median of the values larger than the median.
Values larger than the median are 240, 300, 300, 360, 480, 600.
Median of these values is (300 + 360)/2 = 330.
b. Reported exercise hours per week for the men in the sample ranged from a low of 0 to a high of 600
minutes per week. The median response was 180 minutes (3 hours). About 1/2 of the men (the middle
,Chapter 1 Solutions, 6th ed. Page 4 of 6
half) reported exercising between 37 and 330 minutes (5 and a half hours) per week. About 1/4 said they
exercised less than 37 minutes per week while 1/4 said they exercised more than 330 minutes per week.
1.24 a.
Minutes of Exercise per Week
Median 190
Quartiles 60 285
Extremes 0 450
To determine the summary, first write the responses in order from smallest to largest.
The ordered list of data is:
0, 0, 0, 60, 60, 70, 100, 180, 200, 240, 240, 270, 300, 360, 360, 450
Minimum = 0 min.
Maximum = 450 min.
Median = 190 min. (average of the middle two values in the ordered list, which are 180 and 200)
Lower quartile = 60 min. It is the median of the values that are smaller than the median.
These are 0, 0, 0, 60, 60, 70, 100, 180.
Median of these eight values is (60 + 60)/2 = 60.
Upper quartile = 285 min. It is the median of the values that are larger than the median.
Values larger than the median are 200, 240, 240, 270, 300, 360, 360, 450.
Median of these values is (270 + 300)/2 = 285.
b. Reported exercise hours for the women in the sample ranged from a low of 0 to a high of 450 minutes
per week. The median response was 190 minutes (3 hours and 10 minutes). About 1/2 of the women (the
middle half) reported exercising between 60 and 285 minutes per week. About 1/4 said they exercised
less than 60 minutes per week while 1/4 said they exercised more than 285 minutes per week.
1.25 a. This is an observational study because vegetarians and non-vegetarians are compared and these groups
occur naturally. People were not assigned to treatment groups.
b. Since this is an observational study and not a randomized experiment, we cannot conclude that a
vegetarian diet causes lower death rates from heart attacks and cancer. Other variables not accounted for
may be causing this reduction.
c. This answer will differ for each student. One potential confounding variable is amount of exercise. This
is a confounding variable because it may be that vegetarians also exercise more on average and this led
to lower death rates from heart attacks and cancer.
1.26 Base rates were not given. In this study, a base rate would be the actual rate (risk) of a particular cause of
death for people who are not vegetarians.
1.27 The base rate or baseline risk is missing from the report. You need to know the base rate of cancer of the
rectum for men to decide if the increased risk from drinking beer is large or small.
1.28 a. This was a randomized experiment because volunteers were randomly assigned to wear either a nicotine
patch or a placebo patch.
b. You can conclude that use of nicotine patches leads to a higher success rate for those trying to quit
smoking than use of placebo patches.
c. It was advisable to assign some of the patients to wear a placebo patch because then you can compare
the success rate of those patients to the success rate of the patients wearing nicotine patches. You will
also learn in a future chapter that even though they have no active ingredients, placebos can have a large
psychological effect. Also, presumably people in the experiment want to quit smoking, so some will
succeed regardless of treatment method.
,Chapter 1 Solutions, 6th ed. Page 5 of 6
1.29 For Caution 1: Because the difference given in the previous exercise is a large difference (46% with
nicotine patch and 20% with placebo), it has practical importance as well as statistical significance. For
Caution 2: Because the result is based on a randomized experiment, it is not possible that whether someone
quit or not influenced the type of patch they were assigned.
1.30 The definition is “Statistics is a collection of procedures and principles for gathering data and analyzing
information to help people make decisions when faced with uncertainty.” In the study described in Exercise
1.28, data were gathered and analyzed in order to make a decision about the effectiveness of the nicotine
patch. This information will help individuals decide whether to wear nicotine patches when trying to quit
smoking. Although the observed result was based only on a sample from a larger population, the data
collection and analysis methods make it reasonable to conclude that the patch is more effective than a
placebo for the population represented by this sample.
1.31 Neither caution applies. The magnitude of the difference is given in Case Study 1.6, and considering the
number of men between the ages of 40 and 84 in the U.S. population, the given difference has practical
importance. Because men were randomly assigned to take aspirin (or not) we can conclude that the correct
direction of the cause and effect is that taking aspirin caused the reduction in heart attacks.
1.32 a. The population is all University of California faculty members in 1995.
1 1
b. The margin of error is approximately = = .032 , or roughly 3%.
n 1000
c. It cannot be concluded that a majority of all University faculty favored the criteria. In the sample a slight
majority (52%) was in favor, but the margin of error was about 3%. So, in the population the percent in
favor could possibly be a minority (below 50%). An interval that is 95% certain to contain the percent in
the population is 52% ± 3%, which is 49% to 55%. Because some values within this interval are below
50%, we are not able to rule out the possibility that the percent in favor in the population is a minority.
1 1
1.33 a. The margin of error is about = = .022.
n 2003
b. The sample proportion is 360/2003 = .180. So, the interval is .180 ± .022, which is .158 to .202.
This is sample proportion ± margin of error.
1.34 For this population, the risk of seeing a ghost is 360/2003 = .180.
1.35 a. This is a self-selected (or volunteer) sample.
b. Probably higher, because people who would say they have seen a ghost would be more likely to call the
late-night radio talk show than others. They might even be more likely to be listening to such a show.
1.36 a. Self-selected sample or volunteer sample.
b. The results cannot be extended to any larger population because the sample was not selected to be
representative of any population.
1.37 The term data snooping refers to looking at the data in a variety of ways until something interesting to
report emerges.
1.38 a. This statement has to be based on an observational study. The researchers observed who was breast-fed
and who was not. It would not be possible to randomly assign mothers to breast-feed or not. It would not
be ethical to randomly assign this treatment.
b. The better headline is “Link found between breast-feeding and school performance” because it does not
imply that there is a cause-and-effect conclusion, while the other headline does. Because the study was
observational, a cause-and-effect conclusion cannot be made. It is likely that children who were breast-
fed as infants have other (confounding) factors in their lives that differ from children who were not, and
that may influence school performance. For example, perhaps they are more likely to have mothers who
don’t work, perhaps they are more likely to be first-born children, and so on. We can say that a link was
found, but cannot say that breast-feeding leads to better performance.
1.39 In some situations it is not practical or even possible to conduct a randomized experiment. For example, a
researcher may wish to study whether occupational exposure to asbestos affects the risk of lung disease. It
, Chapter 1 Solutions, 6th ed. Page 6 of 6
would not be possible, or ethical, to assign people to occupations that involve differing amounts of
exposure to asbestos.
1.40 An observational study was done instead of an experiment because the researchers could not assign
individuals either to attend a religious service once a week and pray regularly or to not engage in these
practices.
1.41 The answer will differ for students, but here is an example. Randomly assign volunteers to either eat lots of
chocolate or not eat any chocolate for a period of time, and give them a questionnaire about depression at
the beginning and the end of the time period. Then compare the change in depression scores for the two
groups.
1.42 a. Randomized experiment. People would not take a placebo as part of an observational study.
b. Maybe not, because 20 side effects were tested so one or a few could appear to be a problem
(statistically significant) just by chance even if none of them were a problem. In other words, it is
possible that the observed relationship between taking aspirin and having headaches was a false
positive.
1.43 USA Today made the mistake of making a cause-and-effect conclusion about the relationship between
prayer and blood pressure. This conclusion is not justified because the data were from an observational
study. Specifically, they neglected to consider possible confounding variables like lifestyle choices, social
networks, and health of the people between the two groups. As a result, people may be led to believe that if
they pray more often, they will have lower blood pressure. That conclusion is not justified based on this
observational study.
1.44 Step 1: The investigators asked if taking aspirin reduces the risk of a heart attack and then in the
conclusion, asked to what population the results of this study apply.
Step 2: They collected data from a 5-year randomized experiment, including what treatment each doctor
received (aspirin or placebo), and whether each doctor had a heart attack or not.
Step 3: They summarized the data from Step 2 by categorizing the doctors according to which type of pill
they took (aspirin or placebo). They then counted how many doctors in each treatment group had a
heart attack. Further data analysis included calculating the “Attacks Per 1000 Doctors” for each
group.
Step 4: The investigators proceeded to make conclusions from the data given in Table 1.1. Specifically,
they stated, “The results … support the conclusion that taking aspirin does indeed help reduce the
risk of having a heart attack. The rate of heart attacks in the group taking aspirin was only about
half the rate of heart attacks in the placebo group. In the aspirin group, there were 9.42 heart
attacks per 1000 participating doctors, while in the placebo group, there were 17.13 heart attacks
per 1000 participants.” They then continued to question if there were any other important risk
factors or differences between the two groups.
Step 5: The new knowledge is that this study has provided support for the benefit of aspirin in broader
populations concerning the reduction of the risk in heart attacks. As a result, millions of people
take aspirin with the hope that it will prevent a heart attack.
1.45 a. Categorizing the adults by sex would be more appropriate because cancer type is likely to differ based
on biological sex rather than by the social construct of gender. For instance, biological females cannot
get prostate cancer and biological males cannot get ovarian cancer.
b. Categorizing by gender would be more appropriate because the question of interest is whether there is a
relationship between gender self-identity and opinion on the death penalty.
c. Categorizing by gender would be more appropriate because age at marriage is a social choice, not a
biological trait.
d. Categorizing by sex would be more appropriate because body temperature is likely to differ based on
biological sex and not based on social identity.
Jessica M. Utts
Notes
1- The file is chapter after chapter.
2- We have shown you 10 or five pages.
3- The file contains all Appendix and Excel
sheet if it exists.
4- We have all what you need, we make
update at every time. There are many
new editions waiting you.
5- If you think you purchased the wrong file
You can contact us at every time, we can
replace it with true one.
Our email:
,Mind on Statistics
6th Edition
Complete Solutions Manual
© 2022 Cengage Learning, Inc.
, © 2022 Cengage Learning, Inc. ISBN: 978-133779484-8
Cengage
Unless otherwise noted, all content is © Cengage. 200 Pier 4 Boulevard
Boston, MA 02210
ALL RIGHTS RESERVED. No part of this work covered by the USA
copyright herein may be reproduced or distributed in any form
or by any means, except as permitted by U.S. copyright law, Cengage is a leading provider of customized learning
without the prior written permission of the copyright holder. solutions with employees residing in nearly 40 different
countries and sales in more than 125 countries around
the world. Find your local representative at:
www.cengage.com.
For product information and technology assistance, contact us at
Cengage Customer & Sales Support, To learn more about Cengage platforms and services,
1-800-354-9706 or support.cengage.com. register or access your online learning solution, or
purchase materials for your course, visit
For permission to use material from this text or product, submit www.cengage.com.
all requests online at www.cengage.com/permissions.
NOTE: UNDER NO CIRCUMSTANCES MAY THIS MATERIAL OR ANY PORTION THEREOF BE SOLD, LICENSED, AUCTIONED,
OR OTHERWISE REDISTRIBUTED EXCEPT AS MAY BE PERMITTED BY THE LICENSE TERMS HEREIN.
READ IMPORTANT LICENSE INFORMATION
Dear Professor or Other Supplement Recipient: electronically must be through a password-protected site, with all
copy and download functionality disabled, and accessible solely by
Cengage Learning has provided you with this product (the your students who have purchased the associated textbook for the
“Supplement”) for your review and, to the extent that you adopt Course. You may not sell, license, auction, or otherwise
the associated textbook for use in connection with your course redistribute the Supplement in any form. We ask that you take
(the “Course”), you and your students who purchase the textbook reasonable steps to protect the Supplement from unauthorized
may use the Supplement as described below. Cengage Learning use, reproduction, or distribution. Your use of the Supplement
has established these use limitations in response to concerns indicates your acceptance of the conditions set forth in this
raised by authors, professors, and other users regarding the Agreement. If you do not accept these conditions, you must return
pedagogical problems stemming from unlimited distribution of the Supplement unused within 30 days of receipt.
Supplements.
All rights (including without limitation, copyrights, patents, and
Cengage Learning hereby grants you a nontransferable license to trade secrets) in the Supplement are and will remain the sole and
use the Supplement in connection with the Course, subject to the exclusive property of Cengage Learning and/or its licensors. The
following conditions. The Supplement is for your personal, Supplement is furnished by Cengage Learning on an “as is” basis
noncommercial use only and may not be reproduced, or without any warranties, express or implied. This Agreement will
distributed, except that portions of the Supplement may be be governed by and construed pursuant to the laws of the State
provided to your students in connection with your instruction of the of New York, without regard to such State’s conflict of law rules.
Course, so long as such students are advised that they may not
copy or distribute any portion of the Supplement to any third party. Thank you for your assistance in helping to safeguard the
Test banks and other testing materials may be made available in integrity of the content contained in this Supplement. We trust you
the classroom and collected at the end of each class session, or find the Supplement a useful teaching tool.
posted electronically as described herein. Any material posted
Printed in the United States of America
Print Number: 01 Print Year: 2022
, CHAPTER 1
EXERCISE SOLUTIONS
1.1 a. The fastest speed was 150 miles per hour.
b. The slowest speed driven by a male was 55 miles per hour.
c. 1/4 of the females reported having driven at 95 miles per hour or faster. Notice that
95 mph is the upper quartile for females. By definition, about 1/4 of the values in a dataset are greater
than the upper quartile.
d. 1/2 of the females reported having driven 89 mph or faster. Notice that 89 mph is the median value.
e. 1/2 of 102 = 51 females have driven 89 mph or faster.
Note: For parts (d) and (e) the answer would have to be adjusted if there were any females who reported 89
as their value, but from the data on page 4 we can see that there were not. Because there were no “ties” with
the median, we know that exactly half of the values fall above it and the other half fall below it.
1.2 a. The median height is 65 inches.
b. Range = tallest – shortest = 71 – 59 = 12 inches.
c. The interval from 59 to 63.5 inches contains the shortest 1/4 of these students. This interval is from the
minimum to the lower quartile.
d. The interval from 63.5 to 67.5 inches contains the middle 1/2 of these students. This is an interval from
the lower quartile to the upper quartile.
1.3 a. The observed rate of cervical cancer in Vietnamese American women is 86 per 200,000. This could also
be expressed as 43 per 100,000 or 4.3 per 10,000, and so on. In decimal form, it is .00043.
b. The risk of developing cervical cancer for Vietnamese American women in the next year is 86/200,000
= .00043.
c. The rate of 86 per 200,000 is based on past data and tells us the number of Vietnamese American
women who developed cervical cancer out of a population of 200,000. The risk utilizes the rate from the
past to tell us the future likelihood of cervical cancer in other Vietnamese American women.
1.4 a. The base rate is about 13 in 1000, or about .013.
b. The risk for men who smoke is just over 13 times the rate for non-smokers, or about .169.
1.5 a. All teens in the United States at the time the poll was taken.
b. All teens in the United States who had dated at the time the poll was taken.
1.6 A population is a collection of all individuals of interest while a sample is a subset of the population of
interest, for which measurements are taken in a study. In Case Study 1.6, the population of interest is
probably all men, and possibly women as well. However, the sample consisted of 22,071 male physicians
who volunteered for the study, so the population to which the results apply is all men similar to them.
1.7 a. All adults in the United States at the time the poll was taken.
1
b. = .031 or 3.1%.
1048
c. 34% ± 3.1%, or 30.9% to 37.1%.
1.8 a. The population is probably all Canadians who were eligible to participate in the survey (which is
probably all adults with telephones).
b. There were 2000 people in the sample.
1 1
c. = = .022 or 2.2%.
n 2000
d. In the sample, 16% viewed immigration as having a negative impact. The interval 16% ± 2.2%, or
13.8% to 18.2%, is 95% certain to cover the true percent of Canadians who viewed immigration as
having a negative impact (at the time of the poll).
Page 1 of 6
,Chapter 1 Solutions, 6th ed. Page 2 of 6
1 1
1.9 Solve for n in the equation = .05 = . Answer is n = 400 teenagers.
n 20
1
1.10 Solve for n in the equation = .10. Answer is n = 100.
n
1.11 a. This is an example of a self-selected or volunteer sample. Magazine readers voluntarily responded to the
survey and were not randomly selected.
b. These results may not represent the opinions of all readers of the magazine. The people who respond
probably do so because they feel stronger about the issues (for example, violence on television or
physical discipline) than the readers who do not respond. So, they may be likely to have a generally
different point of view than those who do not respond.
1.12 The exercise did not specify what the survey is about, but no matter what, the survey is based on a
self-selected sample, and people who feel strongly about the issues and/or who have extra time are more
likely to respond. The results will not be representative of all students who use the cafeteria.
1.13 a. Randomized experiment (because students were randomly assigned to the two methods).
b. Observational study (because people cannot be randomly assigned to smoke or not).
c. Observational study (because people cannot be randomly assigned to be a CEO or not).
1.14 a. Randomized experiment, because students were randomly assigned to receive Vitamin C or placebo.
b. Observational study, because the patients were not randomly assigned to do anything. (Note that a
random sample is not the same thing as random assignment.)
c. Randomized experiment, because participants were randomly assigned to meditation or low-fat diet.
1.15 Answers will vary, but one possibility is general level of activity. It is likely to differ for elderly people
who attend church regularly and those who don’t, and it is also likely to affect blood pressure. So it might
partially explain the results of this study.
1.16 a. Number of courses might be a confounding variable. Students taking many courses may sleep less due
to the amount of work involved and may not do as well in school due to the load.
b. Weight is not likely to be a confounding variable. Weight is probably not related either to amount of
sleep or to grades.
c. Hours spent partying might be a confounding variable. Students who party a lot may sleep less, and may
also get lower grades because they’re not studying.
1.17 You would need to know how large the difference in weight loss was for the two groups. If the difference
in weight loss is very small (but not 0), it could be statistically significant, but not have much practical
importance.
1.18 Statistical significance is when there is a relationship or difference that is large enough to be unlikely to
have occurred in the sample if there was no relationship or difference in the population of interest. Practical
significance occurs when the relationship or difference is large enough to be important or meaningful in a
“real world” sense. A result can be statistically significant, but not practically significant. This is especially
likely to occur in studies with very large sample sizes.
1.19 You would want to know how many different relationships were examined. If this result was the only one
that was statistically significant out of many examined, it could easily be a false positive.
1.20 A false positive occurs when a relationship or difference is said to be statistically significant based on
examining information from a sample, but in fact there is no relationship or difference in the population.
1.21 The placebo group estimates the baseline rate of heart attacks for men not taking aspirin. So, the estimated
baseline rate of heart attacks is 189/11,034, which is about 17 heart attacks per 1000 men or 17/1000. (See
Table 1.1 for the data.)
1.22 a. The amount of exercise per week is similar for men and women except that there are a few high values
for the men. The dotplot follows.
,Chapter 1 Solutions, 6th ed. Page 3 of 6
Figure for Exercise 1.22a
Gender
m
w
0 100 200 300 400 500 600
Exercise per week (minutes)
b. Women, median = 190 min. Men, median = 180 min.
To find the median, put the data in order first.
For women, the ordered list of data is:
0, 0, 0, 60, 60, 70, 100, 180, 200, 240, 240, 270, 300, 360, 360, 450
The number of women is even (16), so the median is the average of the middle two values in the ordered
list. These middle two values, underlined and bold in the above list, are 180 and 200 and their average is
190.
For men, the ordered list of data is:
0, 0, 14, 60, 90, 120, 180, 240, 300, 300, 360, 480, 600
The number of men is odd (13), so the median is the middle value in the ordered data. This value,
underlined and bold in the list above, is 180.
c. Although the median response is different for women and men, the difference is only 10 minutes. The
weekly amount of exercise is about the same for the samples of women and men.
1.23 a.
Minutes of Exercise per Week
Median 180
Quartiles 37 330
Extremes 0 600
To determine the summary, first write the responses in order from smallest to largest.
The ordered list of data is:
0, 0, 14, 60, 90, 120, 180, 240, 300, 300, 360, 480, 600
Minimum = 0 min.
Maximum = 600 min.
Median = 180 min. (middle value in the ordered list)
Lower quartile = 42 min. It is the median of the values smaller than the median.
These are 0, 0, 14, 60, 90, 120.
Median of these six values is (14 + 70)/2 = 42.
Upper quartile = 330 min. It is the median of the values larger than the median.
Values larger than the median are 240, 300, 300, 360, 480, 600.
Median of these values is (300 + 360)/2 = 330.
b. Reported exercise hours per week for the men in the sample ranged from a low of 0 to a high of 600
minutes per week. The median response was 180 minutes (3 hours). About 1/2 of the men (the middle
,Chapter 1 Solutions, 6th ed. Page 4 of 6
half) reported exercising between 37 and 330 minutes (5 and a half hours) per week. About 1/4 said they
exercised less than 37 minutes per week while 1/4 said they exercised more than 330 minutes per week.
1.24 a.
Minutes of Exercise per Week
Median 190
Quartiles 60 285
Extremes 0 450
To determine the summary, first write the responses in order from smallest to largest.
The ordered list of data is:
0, 0, 0, 60, 60, 70, 100, 180, 200, 240, 240, 270, 300, 360, 360, 450
Minimum = 0 min.
Maximum = 450 min.
Median = 190 min. (average of the middle two values in the ordered list, which are 180 and 200)
Lower quartile = 60 min. It is the median of the values that are smaller than the median.
These are 0, 0, 0, 60, 60, 70, 100, 180.
Median of these eight values is (60 + 60)/2 = 60.
Upper quartile = 285 min. It is the median of the values that are larger than the median.
Values larger than the median are 200, 240, 240, 270, 300, 360, 360, 450.
Median of these values is (270 + 300)/2 = 285.
b. Reported exercise hours for the women in the sample ranged from a low of 0 to a high of 450 minutes
per week. The median response was 190 minutes (3 hours and 10 minutes). About 1/2 of the women (the
middle half) reported exercising between 60 and 285 minutes per week. About 1/4 said they exercised
less than 60 minutes per week while 1/4 said they exercised more than 285 minutes per week.
1.25 a. This is an observational study because vegetarians and non-vegetarians are compared and these groups
occur naturally. People were not assigned to treatment groups.
b. Since this is an observational study and not a randomized experiment, we cannot conclude that a
vegetarian diet causes lower death rates from heart attacks and cancer. Other variables not accounted for
may be causing this reduction.
c. This answer will differ for each student. One potential confounding variable is amount of exercise. This
is a confounding variable because it may be that vegetarians also exercise more on average and this led
to lower death rates from heart attacks and cancer.
1.26 Base rates were not given. In this study, a base rate would be the actual rate (risk) of a particular cause of
death for people who are not vegetarians.
1.27 The base rate or baseline risk is missing from the report. You need to know the base rate of cancer of the
rectum for men to decide if the increased risk from drinking beer is large or small.
1.28 a. This was a randomized experiment because volunteers were randomly assigned to wear either a nicotine
patch or a placebo patch.
b. You can conclude that use of nicotine patches leads to a higher success rate for those trying to quit
smoking than use of placebo patches.
c. It was advisable to assign some of the patients to wear a placebo patch because then you can compare
the success rate of those patients to the success rate of the patients wearing nicotine patches. You will
also learn in a future chapter that even though they have no active ingredients, placebos can have a large
psychological effect. Also, presumably people in the experiment want to quit smoking, so some will
succeed regardless of treatment method.
,Chapter 1 Solutions, 6th ed. Page 5 of 6
1.29 For Caution 1: Because the difference given in the previous exercise is a large difference (46% with
nicotine patch and 20% with placebo), it has practical importance as well as statistical significance. For
Caution 2: Because the result is based on a randomized experiment, it is not possible that whether someone
quit or not influenced the type of patch they were assigned.
1.30 The definition is “Statistics is a collection of procedures and principles for gathering data and analyzing
information to help people make decisions when faced with uncertainty.” In the study described in Exercise
1.28, data were gathered and analyzed in order to make a decision about the effectiveness of the nicotine
patch. This information will help individuals decide whether to wear nicotine patches when trying to quit
smoking. Although the observed result was based only on a sample from a larger population, the data
collection and analysis methods make it reasonable to conclude that the patch is more effective than a
placebo for the population represented by this sample.
1.31 Neither caution applies. The magnitude of the difference is given in Case Study 1.6, and considering the
number of men between the ages of 40 and 84 in the U.S. population, the given difference has practical
importance. Because men were randomly assigned to take aspirin (or not) we can conclude that the correct
direction of the cause and effect is that taking aspirin caused the reduction in heart attacks.
1.32 a. The population is all University of California faculty members in 1995.
1 1
b. The margin of error is approximately = = .032 , or roughly 3%.
n 1000
c. It cannot be concluded that a majority of all University faculty favored the criteria. In the sample a slight
majority (52%) was in favor, but the margin of error was about 3%. So, in the population the percent in
favor could possibly be a minority (below 50%). An interval that is 95% certain to contain the percent in
the population is 52% ± 3%, which is 49% to 55%. Because some values within this interval are below
50%, we are not able to rule out the possibility that the percent in favor in the population is a minority.
1 1
1.33 a. The margin of error is about = = .022.
n 2003
b. The sample proportion is 360/2003 = .180. So, the interval is .180 ± .022, which is .158 to .202.
This is sample proportion ± margin of error.
1.34 For this population, the risk of seeing a ghost is 360/2003 = .180.
1.35 a. This is a self-selected (or volunteer) sample.
b. Probably higher, because people who would say they have seen a ghost would be more likely to call the
late-night radio talk show than others. They might even be more likely to be listening to such a show.
1.36 a. Self-selected sample or volunteer sample.
b. The results cannot be extended to any larger population because the sample was not selected to be
representative of any population.
1.37 The term data snooping refers to looking at the data in a variety of ways until something interesting to
report emerges.
1.38 a. This statement has to be based on an observational study. The researchers observed who was breast-fed
and who was not. It would not be possible to randomly assign mothers to breast-feed or not. It would not
be ethical to randomly assign this treatment.
b. The better headline is “Link found between breast-feeding and school performance” because it does not
imply that there is a cause-and-effect conclusion, while the other headline does. Because the study was
observational, a cause-and-effect conclusion cannot be made. It is likely that children who were breast-
fed as infants have other (confounding) factors in their lives that differ from children who were not, and
that may influence school performance. For example, perhaps they are more likely to have mothers who
don’t work, perhaps they are more likely to be first-born children, and so on. We can say that a link was
found, but cannot say that breast-feeding leads to better performance.
1.39 In some situations it is not practical or even possible to conduct a randomized experiment. For example, a
researcher may wish to study whether occupational exposure to asbestos affects the risk of lung disease. It
, Chapter 1 Solutions, 6th ed. Page 6 of 6
would not be possible, or ethical, to assign people to occupations that involve differing amounts of
exposure to asbestos.
1.40 An observational study was done instead of an experiment because the researchers could not assign
individuals either to attend a religious service once a week and pray regularly or to not engage in these
practices.
1.41 The answer will differ for students, but here is an example. Randomly assign volunteers to either eat lots of
chocolate or not eat any chocolate for a period of time, and give them a questionnaire about depression at
the beginning and the end of the time period. Then compare the change in depression scores for the two
groups.
1.42 a. Randomized experiment. People would not take a placebo as part of an observational study.
b. Maybe not, because 20 side effects were tested so one or a few could appear to be a problem
(statistically significant) just by chance even if none of them were a problem. In other words, it is
possible that the observed relationship between taking aspirin and having headaches was a false
positive.
1.43 USA Today made the mistake of making a cause-and-effect conclusion about the relationship between
prayer and blood pressure. This conclusion is not justified because the data were from an observational
study. Specifically, they neglected to consider possible confounding variables like lifestyle choices, social
networks, and health of the people between the two groups. As a result, people may be led to believe that if
they pray more often, they will have lower blood pressure. That conclusion is not justified based on this
observational study.
1.44 Step 1: The investigators asked if taking aspirin reduces the risk of a heart attack and then in the
conclusion, asked to what population the results of this study apply.
Step 2: They collected data from a 5-year randomized experiment, including what treatment each doctor
received (aspirin or placebo), and whether each doctor had a heart attack or not.
Step 3: They summarized the data from Step 2 by categorizing the doctors according to which type of pill
they took (aspirin or placebo). They then counted how many doctors in each treatment group had a
heart attack. Further data analysis included calculating the “Attacks Per 1000 Doctors” for each
group.
Step 4: The investigators proceeded to make conclusions from the data given in Table 1.1. Specifically,
they stated, “The results … support the conclusion that taking aspirin does indeed help reduce the
risk of having a heart attack. The rate of heart attacks in the group taking aspirin was only about
half the rate of heart attacks in the placebo group. In the aspirin group, there were 9.42 heart
attacks per 1000 participating doctors, while in the placebo group, there were 17.13 heart attacks
per 1000 participants.” They then continued to question if there were any other important risk
factors or differences between the two groups.
Step 5: The new knowledge is that this study has provided support for the benefit of aspirin in broader
populations concerning the reduction of the risk in heart attacks. As a result, millions of people
take aspirin with the hope that it will prevent a heart attack.
1.45 a. Categorizing the adults by sex would be more appropriate because cancer type is likely to differ based
on biological sex rather than by the social construct of gender. For instance, biological females cannot
get prostate cancer and biological males cannot get ovarian cancer.
b. Categorizing by gender would be more appropriate because the question of interest is whether there is a
relationship between gender self-identity and opinion on the death penalty.
c. Categorizing by gender would be more appropriate because age at marriage is a social choice, not a
biological trait.
d. Categorizing by sex would be more appropriate because body temperature is likely to differ based on
biological sex and not based on social identity.
