, SOLUTION MANUAL FOR An Introduction to
Statistics and Data Analysis Using Stata® From
Research Design to Final Report Second Edition
Lisa Daniels
Notes
1- The file is chapter after chapter.
2- We have shown you few pages sample.
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:
, Full Homework Answer Key
for the textbook
An Introduction to Statistics and Data Analysis Using Stata:
From Research Design to Final Report
Second Edition
by Lisa Daniels and Nicholas W. Minot
We’ve done our very best to make this answer key for the second edition as error-free as
possible. But if a mistake slipped past us, we’d love your help spotting it. Please let us know at
(Lisa Daniels) or (Nick Minot). Thank you!
Chapter 1: The Research Process
1. Read the article “Prevalence and Motives for Illicit Use of Prescription Stimulants in an
Undergraduate Sample” by Teter, et al. As you read the article, answer the questions
below, which are based on guidelines offered by Greenlaw(2009).
a) What question or questions are the authors asking?
What is the prevalence for use of prescription stimulants and alcohol and other drugs
What are the motives for using drugs?
Is there an association between motives for stimulant use and alcohol and other drug use
b) Describe the theoretical approach that the authors use to develop their research
question.
The authors do not directly identify a theoretical approach. They do offer possible reasons for
misuse of stimulants. They also point out that the literature indicates that the misuse of
stimulants helps to obtain “other forms of reinforcement” such as “social interactions, partying,
work, or studying.”
c) What answers do the authors propose?
8.1% reported lifetime use
5.4% reported past-year use
Motives – help with concentration, increase alertness, provide a high
d) In what ways does the current study improve over previous research according to
the authors of the article? In other words, what gaps do the authors identify in the
current literature?
-only 12 published reports
-motives for drinking examined, but not for stimulants
e) What method do the authors use to answer their questions?
1
,Web-based survey to a random sample of 9,161 undergrads
f) What limitations do the authors identify in their own study?
Gaps:
-single campus
-predominantly white population at a large public university
-nonresponses
-didn’t ask about quantity of drugs used per occasion
-didn’t ask about route of administration
g) What suggestions do the authors have for follow-up research that should be done?
Future research:
-need for qualitative research to assess motives in various populations – urban residents, not
attending college, divers racial backgrounds
-refinement of measure of motives on Liert-type scale
-comparing motives for different forms of route (intranasal, intravenous
-long-term morbidity and mortality associated with stimulants
-longitudinal studies – over time the effect
-who writes the prescriptions
-study on quantity of scrips being written
2. Choose a general area of research that interests you. This could be sports, cancer,
poverty, social media usage, gaming, etc. Use the techniques identified in Section 1.2 to
narrow your focus as you begin perusing the literature and using forward and
backward searching for articles of particular interest to you. Once you have done the
initial reading, you should then develop a tentative research question and identify five
articles that are most closely related to your question. For each of the five articles,
answer the following questions:
a) What question or questions are the authors asking?
b) Describe the theoretical approach that the authors use to develop their research
question.
c) What is the hypothesis that the authors propose?
d) What answers do the authors propose?
e) In what ways does the current study improve over previous research according to
the authors of the article? In other words, what gaps do the authors identify in the
current literature?
f) What method do the authors use to answer their questions?
g) What limitations do the authors identify in their study?
h) What suggestions do the authors have for follow-up research that should be done?
Answers to question 2 will be unique to each student.
2
,Chapter 2: Sampling Techniques
1. Suppose you have a sampling frame of 1,200 hardware stores in a state, numbered from
1 to 1,200. Describe how you would select a systematic random sample of 100 stores for
a survey. Given an example of what the sample might look like, showing the store
numbers of the first five selected stores.
The first step is to select randomly from among the first N/n stores, that is among the first
1200/100=12 stores. The rest of the sample is drawn by selecting every N/n=12 stores starting
with the initial one. The first store should be between 1 and 12, while the remaining stores
should have an interval of 12 between each pair. For example, if store #6 is selected randomly as
the first store, the other stores would be 18, 30, 42, 54, 66, 78, and so on.
2. Give three possible reasons why one might want to use a multi-stage random sample
rather than a single-stage random sample.
First, multi-stage random sample could be useful if the full sampling frame is not available. For
example, if household lists are only available at the county level, it would be easier to select
counties and then individual units for each county. Second, multi-stage sampling may be used to
ensure that the sample is spread proportionally across categories, such as states. Third, multi-
stage sampling may be used to reduce the cost of data collection because the sample is clustered.
3. Describe the general circumstances under which it would be useful to apply area
sampling to select units to interview. Give an example of a situation in which area
sampling would be useful.
Often a list of sampling units is not available. For this reason, you would need to use an area
sampling frame. For example, supposed you wanted to determine the number of people in your
country who do not eat food with gluten. You don’t have a list of people in your county. You
would then use a map of the county divided into areas with 10 house each. You could then select
one house out of each 10 houses to send your survey.
4. You have been hired to design a survey of political opinions in 10 “swing” states, but
you need to have a large enough sample (say, 400 respondents per state) to generate
reliable results for each state. What type of sampling method do you need to use?
To ensure a sufficient sample size in each state, it is necessary to use a stratified random sample.
The sample will be stratified at the state level, with at least 400 respondents per state that is in
the sample.
5. Assuming you have a list of all households in each state and can use simple random
sampling in each, how would you calculate the sampling weight for each household in
the survey?
The sampling weight for each state would be the population of that state divided by the number
of households that are included in your survey in that state.
3
,6. There are 20,000 people in the country of Wakanda. Most of the population, 17,500,
lives in urban areas and the rest live in rural areas. If you drew a sample of 250 people
from urban areas and 250 people from rural areas, what would be the weights for
urban and rural areas?
17,500
𝑈𝑟𝑏𝑎𝑛 𝑤𝑒𝑖𝑔ℎ𝑡 = = 70
250
2,500
𝑅𝑢𝑟𝑎𝑙 𝑤𝑒𝑖𝑔ℎ𝑡 = = 10
250
7. Suppose that we develop a multi-stage sampling design and choose 5 states (out of 50)
and 3 counties within each state, and 300 households in each county. In the state of
Pennsylvania where there are 67 counties, we randomly select the following three
counties:
County Population
Montgomery 819,000
Bucks 630,000
Allegheny 1,200,000
What is the weight for one household in each of these three counties? Round your
answer to the nearest whole number.
This is a three-stage random sample, so the weight is calculated as the product of three ratios,
one for each stage: states, counties, and households. Note that the population of each county
must be divided by three to get the number of households per county.
50 67 819,000
𝑀𝑜𝑛𝑡𝑔𝑜𝑚𝑒𝑟𝑦 𝐶𝑜𝑢𝑛𝑡𝑦 𝑊𝑒𝑖𝑔ℎ𝑡 = ∗ ∗ = 609,700
5 3 300
50 67 630,000
𝐵𝑢𝑐𝑘𝑠 𝐶𝑜𝑢𝑛𝑡𝑦 𝑊𝑒𝑖𝑔ℎ𝑡 = ∗ ∗ = 469,000
5 3 300
50 67 1,200,000
𝐴𝑙𝑙𝑒𝑔ℎ𝑒𝑛𝑦 𝐶𝑜𝑢𝑛𝑡𝑦 𝑊𝑒𝑖𝑔ℎ𝑡 = ∗ ∗ = 893,333
5 3 300
4
,Chapter 3: Questionnaire Design
1. Which of the following questions generate responses that are continuous variable
responses and which ones generate responses that are categorical variables?
a) How old are you?
Continuous
b) What is the highest educational degree you have completed?
Categorical
c) How many years of education do you have?
Continuous
d) What state were you born in?
Categorical
e) Are both of your parents still alive?
Categorical
f) How many times per week do you exercise?
Continuous
g) What is your weight in pounds?
Continuous
2. Identify the hidden assumption(s) or the flaw(s) in the following questions.
a) How much do you earn at your job?
This needs to specify the time unit – per hour, per week, per month.
It also assumes that the respondent has only one job.
b) What is the age of your oldest child?
This assumes that the respondent has children.
c) In light of his ineffectiveness, do you agree that the governor should not be
reelected?
This is a loaded question that assumes that the respondent thinks that the governor is ineffective.
d) How old (in years) is your car?
It assumes that the responded owns a car.
It assumes that the respondent owns just one car.
e) How frequently do you go shopping?
This needs a time unit – per week, per month
It is not specific about the type of shopping – grocery, clothes, or something else?
It doesn’t specify whether the question refers to the respondent alone or the respondent’s household.
5
,It assumes that the respondent goes shopping, rather than someone else in the household.
f) Do you think the town firefighters should be full-time workers and paid more?
This assume that all of the firefighters are part-time.
This is double-barreled in that you are asking two different questions.
3. What are skip patterns in a questionnaire, and what purpose do they serve?
If a respondent answers a question as yes or no, it might affect whether they should see the next round of
questions. For example, if you are asked how many hours a week do you drive a car. They should first
be asked if they drive a car. If not, then they should not see the next question about the hours that they
drive. Respondents should only see questions that are relevant to them.
4. Give an example of social desirability bias and how it might affect the accuracy of
results in a survey.
Individuals want to present a good version of themselves. So when asked if they would help a homeless
person, they may say “yes” even if it is not true. Similarly, they may exaggerate the amount of money
they give to charity or lie about the number of friends they have.
5. How would you use skip patterns to design questions to gather information about the
car payments of a sample of respondents?
Do you own a car? (if no, skip the next two questions)
Did you take out a loan that you are still paying off? (if no, skip the next question)
How much are the monthly payments for your car?
6. What are the three main principles for research on human subjects used by IRBs to
approve research? Given an example of a violation of each principle.
Informed consent – A survey is sent out that does not specifically ask if the respondent consents to
the survey.
Confidentiality – The data related to individuals are shared with the public
Costs and risks to respondents are justified by benefits of the survey – The risks for respondents are
far greater than the benefits. For example, a new drug is being tested on respondents, but the risks are
high due to limited testing.
6
,Chapter 4: An Introduction to Stata
1. Ten college students were asked four questions about their streaming habits:
Which streaming service do you use most often to watch television shows and movies?
How many hours a week do you spend watching series or movies?
How often do you binge watch shows (watching more than three episodes of the same
show in a row)? They could choose from (a) not at all, (b) sometimes—one to three times
per week, and (c) frequently— more than three times per week.
Gender: How do you identify? (a) female, (b) male, (c) nonbinary (d) other
a) Based on their responses that are in the table below, Enter the data for each of the
four variables. For the three categorical variable (streaming service, frequency of
binging, and gender), create a numeric code for each response. For example, for
streaming service, 1 = Amazon Prime, 2 = Hulu Plus, 3 = HBO, and 4 = Netflix.
HINT: A common mistake is to type the variable names into the first row of the
spreadsheet. Do not type variable names into the data editor spreadsheet. You
should use Stata code to rename the variables, which will appear at the top of each
column in the data editor spreadsheet.
b) Once you have entered the data use a do-file to rename each variable.
See Stata code below.
c) Give each variable a variable label.
See Stata code below.
d) Give each numeric code a value label.
See Stata code below.
e) Save your data file (you will use this again in a later chapter).
7
, Student Responses to Survey About TV and Movie Viewing Habits
Student TV Source Hours per Binge Frequency Sex at birth
Week
1 Hulu Plus 14 Not at all Male
2 Amazon Prime 18 Sometimes Female
3 Hulu Plus 20 Frequently Female
4 Netflix 5 Frequently Nonbinary
5 Netflix 12 Frequently Male
6 HBO 10 Not at all Female
7 HBO 8 Frequently Female
8 HBO 7 Sometimes Other
9 Amazon Prime 24 Frequently Male
10 Hulu Plus 30 Sometimes Female
8
Statistics and Data Analysis Using Stata® From
Research Design to Final Report Second Edition
Lisa Daniels
Notes
1- The file is chapter after chapter.
2- We have shown you few pages sample.
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:
, Full Homework Answer Key
for the textbook
An Introduction to Statistics and Data Analysis Using Stata:
From Research Design to Final Report
Second Edition
by Lisa Daniels and Nicholas W. Minot
We’ve done our very best to make this answer key for the second edition as error-free as
possible. But if a mistake slipped past us, we’d love your help spotting it. Please let us know at
(Lisa Daniels) or (Nick Minot). Thank you!
Chapter 1: The Research Process
1. Read the article “Prevalence and Motives for Illicit Use of Prescription Stimulants in an
Undergraduate Sample” by Teter, et al. As you read the article, answer the questions
below, which are based on guidelines offered by Greenlaw(2009).
a) What question or questions are the authors asking?
What is the prevalence for use of prescription stimulants and alcohol and other drugs
What are the motives for using drugs?
Is there an association between motives for stimulant use and alcohol and other drug use
b) Describe the theoretical approach that the authors use to develop their research
question.
The authors do not directly identify a theoretical approach. They do offer possible reasons for
misuse of stimulants. They also point out that the literature indicates that the misuse of
stimulants helps to obtain “other forms of reinforcement” such as “social interactions, partying,
work, or studying.”
c) What answers do the authors propose?
8.1% reported lifetime use
5.4% reported past-year use
Motives – help with concentration, increase alertness, provide a high
d) In what ways does the current study improve over previous research according to
the authors of the article? In other words, what gaps do the authors identify in the
current literature?
-only 12 published reports
-motives for drinking examined, but not for stimulants
e) What method do the authors use to answer their questions?
1
,Web-based survey to a random sample of 9,161 undergrads
f) What limitations do the authors identify in their own study?
Gaps:
-single campus
-predominantly white population at a large public university
-nonresponses
-didn’t ask about quantity of drugs used per occasion
-didn’t ask about route of administration
g) What suggestions do the authors have for follow-up research that should be done?
Future research:
-need for qualitative research to assess motives in various populations – urban residents, not
attending college, divers racial backgrounds
-refinement of measure of motives on Liert-type scale
-comparing motives for different forms of route (intranasal, intravenous
-long-term morbidity and mortality associated with stimulants
-longitudinal studies – over time the effect
-who writes the prescriptions
-study on quantity of scrips being written
2. Choose a general area of research that interests you. This could be sports, cancer,
poverty, social media usage, gaming, etc. Use the techniques identified in Section 1.2 to
narrow your focus as you begin perusing the literature and using forward and
backward searching for articles of particular interest to you. Once you have done the
initial reading, you should then develop a tentative research question and identify five
articles that are most closely related to your question. For each of the five articles,
answer the following questions:
a) What question or questions are the authors asking?
b) Describe the theoretical approach that the authors use to develop their research
question.
c) What is the hypothesis that the authors propose?
d) What answers do the authors propose?
e) In what ways does the current study improve over previous research according to
the authors of the article? In other words, what gaps do the authors identify in the
current literature?
f) What method do the authors use to answer their questions?
g) What limitations do the authors identify in their study?
h) What suggestions do the authors have for follow-up research that should be done?
Answers to question 2 will be unique to each student.
2
,Chapter 2: Sampling Techniques
1. Suppose you have a sampling frame of 1,200 hardware stores in a state, numbered from
1 to 1,200. Describe how you would select a systematic random sample of 100 stores for
a survey. Given an example of what the sample might look like, showing the store
numbers of the first five selected stores.
The first step is to select randomly from among the first N/n stores, that is among the first
1200/100=12 stores. The rest of the sample is drawn by selecting every N/n=12 stores starting
with the initial one. The first store should be between 1 and 12, while the remaining stores
should have an interval of 12 between each pair. For example, if store #6 is selected randomly as
the first store, the other stores would be 18, 30, 42, 54, 66, 78, and so on.
2. Give three possible reasons why one might want to use a multi-stage random sample
rather than a single-stage random sample.
First, multi-stage random sample could be useful if the full sampling frame is not available. For
example, if household lists are only available at the county level, it would be easier to select
counties and then individual units for each county. Second, multi-stage sampling may be used to
ensure that the sample is spread proportionally across categories, such as states. Third, multi-
stage sampling may be used to reduce the cost of data collection because the sample is clustered.
3. Describe the general circumstances under which it would be useful to apply area
sampling to select units to interview. Give an example of a situation in which area
sampling would be useful.
Often a list of sampling units is not available. For this reason, you would need to use an area
sampling frame. For example, supposed you wanted to determine the number of people in your
country who do not eat food with gluten. You don’t have a list of people in your county. You
would then use a map of the county divided into areas with 10 house each. You could then select
one house out of each 10 houses to send your survey.
4. You have been hired to design a survey of political opinions in 10 “swing” states, but
you need to have a large enough sample (say, 400 respondents per state) to generate
reliable results for each state. What type of sampling method do you need to use?
To ensure a sufficient sample size in each state, it is necessary to use a stratified random sample.
The sample will be stratified at the state level, with at least 400 respondents per state that is in
the sample.
5. Assuming you have a list of all households in each state and can use simple random
sampling in each, how would you calculate the sampling weight for each household in
the survey?
The sampling weight for each state would be the population of that state divided by the number
of households that are included in your survey in that state.
3
,6. There are 20,000 people in the country of Wakanda. Most of the population, 17,500,
lives in urban areas and the rest live in rural areas. If you drew a sample of 250 people
from urban areas and 250 people from rural areas, what would be the weights for
urban and rural areas?
17,500
𝑈𝑟𝑏𝑎𝑛 𝑤𝑒𝑖𝑔ℎ𝑡 = = 70
250
2,500
𝑅𝑢𝑟𝑎𝑙 𝑤𝑒𝑖𝑔ℎ𝑡 = = 10
250
7. Suppose that we develop a multi-stage sampling design and choose 5 states (out of 50)
and 3 counties within each state, and 300 households in each county. In the state of
Pennsylvania where there are 67 counties, we randomly select the following three
counties:
County Population
Montgomery 819,000
Bucks 630,000
Allegheny 1,200,000
What is the weight for one household in each of these three counties? Round your
answer to the nearest whole number.
This is a three-stage random sample, so the weight is calculated as the product of three ratios,
one for each stage: states, counties, and households. Note that the population of each county
must be divided by three to get the number of households per county.
50 67 819,000
𝑀𝑜𝑛𝑡𝑔𝑜𝑚𝑒𝑟𝑦 𝐶𝑜𝑢𝑛𝑡𝑦 𝑊𝑒𝑖𝑔ℎ𝑡 = ∗ ∗ = 609,700
5 3 300
50 67 630,000
𝐵𝑢𝑐𝑘𝑠 𝐶𝑜𝑢𝑛𝑡𝑦 𝑊𝑒𝑖𝑔ℎ𝑡 = ∗ ∗ = 469,000
5 3 300
50 67 1,200,000
𝐴𝑙𝑙𝑒𝑔ℎ𝑒𝑛𝑦 𝐶𝑜𝑢𝑛𝑡𝑦 𝑊𝑒𝑖𝑔ℎ𝑡 = ∗ ∗ = 893,333
5 3 300
4
,Chapter 3: Questionnaire Design
1. Which of the following questions generate responses that are continuous variable
responses and which ones generate responses that are categorical variables?
a) How old are you?
Continuous
b) What is the highest educational degree you have completed?
Categorical
c) How many years of education do you have?
Continuous
d) What state were you born in?
Categorical
e) Are both of your parents still alive?
Categorical
f) How many times per week do you exercise?
Continuous
g) What is your weight in pounds?
Continuous
2. Identify the hidden assumption(s) or the flaw(s) in the following questions.
a) How much do you earn at your job?
This needs to specify the time unit – per hour, per week, per month.
It also assumes that the respondent has only one job.
b) What is the age of your oldest child?
This assumes that the respondent has children.
c) In light of his ineffectiveness, do you agree that the governor should not be
reelected?
This is a loaded question that assumes that the respondent thinks that the governor is ineffective.
d) How old (in years) is your car?
It assumes that the responded owns a car.
It assumes that the respondent owns just one car.
e) How frequently do you go shopping?
This needs a time unit – per week, per month
It is not specific about the type of shopping – grocery, clothes, or something else?
It doesn’t specify whether the question refers to the respondent alone or the respondent’s household.
5
,It assumes that the respondent goes shopping, rather than someone else in the household.
f) Do you think the town firefighters should be full-time workers and paid more?
This assume that all of the firefighters are part-time.
This is double-barreled in that you are asking two different questions.
3. What are skip patterns in a questionnaire, and what purpose do they serve?
If a respondent answers a question as yes or no, it might affect whether they should see the next round of
questions. For example, if you are asked how many hours a week do you drive a car. They should first
be asked if they drive a car. If not, then they should not see the next question about the hours that they
drive. Respondents should only see questions that are relevant to them.
4. Give an example of social desirability bias and how it might affect the accuracy of
results in a survey.
Individuals want to present a good version of themselves. So when asked if they would help a homeless
person, they may say “yes” even if it is not true. Similarly, they may exaggerate the amount of money
they give to charity or lie about the number of friends they have.
5. How would you use skip patterns to design questions to gather information about the
car payments of a sample of respondents?
Do you own a car? (if no, skip the next two questions)
Did you take out a loan that you are still paying off? (if no, skip the next question)
How much are the monthly payments for your car?
6. What are the three main principles for research on human subjects used by IRBs to
approve research? Given an example of a violation of each principle.
Informed consent – A survey is sent out that does not specifically ask if the respondent consents to
the survey.
Confidentiality – The data related to individuals are shared with the public
Costs and risks to respondents are justified by benefits of the survey – The risks for respondents are
far greater than the benefits. For example, a new drug is being tested on respondents, but the risks are
high due to limited testing.
6
,Chapter 4: An Introduction to Stata
1. Ten college students were asked four questions about their streaming habits:
Which streaming service do you use most often to watch television shows and movies?
How many hours a week do you spend watching series or movies?
How often do you binge watch shows (watching more than three episodes of the same
show in a row)? They could choose from (a) not at all, (b) sometimes—one to three times
per week, and (c) frequently— more than three times per week.
Gender: How do you identify? (a) female, (b) male, (c) nonbinary (d) other
a) Based on their responses that are in the table below, Enter the data for each of the
four variables. For the three categorical variable (streaming service, frequency of
binging, and gender), create a numeric code for each response. For example, for
streaming service, 1 = Amazon Prime, 2 = Hulu Plus, 3 = HBO, and 4 = Netflix.
HINT: A common mistake is to type the variable names into the first row of the
spreadsheet. Do not type variable names into the data editor spreadsheet. You
should use Stata code to rename the variables, which will appear at the top of each
column in the data editor spreadsheet.
b) Once you have entered the data use a do-file to rename each variable.
See Stata code below.
c) Give each variable a variable label.
See Stata code below.
d) Give each numeric code a value label.
See Stata code below.
e) Save your data file (you will use this again in a later chapter).
7
, Student Responses to Survey About TV and Movie Viewing Habits
Student TV Source Hours per Binge Frequency Sex at birth
Week
1 Hulu Plus 14 Not at all Male
2 Amazon Prime 18 Sometimes Female
3 Hulu Plus 20 Frequently Female
4 Netflix 5 Frequently Nonbinary
5 Netflix 12 Frequently Male
6 HBO 10 Not at all Female
7 HBO 8 Frequently Female
8 HBO 7 Sometimes Other
9 Amazon Prime 24 Frequently Male
10 Hulu Plus 30 Sometimes Female
8
