Chapter 5 Organization of Data for Analysis

Chapter 5 Prerequisite Skills

Chapter 5 Prerequisite Skills Question 1 Page 194

a) This is a line graph, which uses line segments to connect data points to show changes in data
over time.

b) This is a double bar graph, which uses horizontal or vertical bars whose lengths are
proportional to the quantities to represent two sets of data.

c) This is a scatterplot, which uses ordered pairs to see relationships between the two variables.

d) This is a circle graph, which uses sectors of a circle that are proportional to the quantities to
represent data.

e) This is a bar graph, which uses horizontal or vertical bars whose lengths are proportional to
the quantities to represent data.

f) This is a histogram, which is a type of bar graph in which the bars are connected and are
proportional to the frequencies for a range of values of the variable.

Chapter 5 Prerequisite Skills Question 2 Page 194

a) The histogram shows intervals of 10 to 14, 15 to 19, and 20 to 24.

b) The approximate cost to deliver mail to a centralized point is a little over $100 per address.

c) Incidents of crime seem to increase with an increase in officers.

d) The most popular game console is the Xbox 360 at 30%.

e) Since the only time period that does not show an increase is the month of November, that
must be when students had their cell phones taken away.

f) Breaking Dawn - Part 2 made the most money on its opening weekend.

Chapter 5 Prerequisite Skills Question 3 Page 194

A bar graph shows the frequencies of the values of the variable with disconnected bars and can be
used to represent non-numeric or unordered data. A histogram uses connected bars and represents
a continuous range of values in numeric order.

MHR  Data Management 12 Solutions 1

Chapter 5 Prerequisite Skills Question 4 Page 194

Chapter 5 Prerequisite Skills Question 5 Page 195

Chapter 5 Prerequisite Skills Question 6 Page 195

2 MHR  Data Management 12 Solutions

Chapter 5 Prerequisite Skills Question 7 Page 195

Chapter 5 Section 1 Data Concepts and Graphical Summaries

Chapter 5 Section 1 Example 1 Your Turn Page 197

Answers may vary.

Climate change: Ninety-seven percent of climate scientists agree that climate-warming trends
over the past century are very likely due to human activities. Others argue that the observed
warming is more likely attributable to natural causes than to human activities.

Chapter 5 Section 1 Example 2 Your Turn Page 199

These are categorical, ordinal data since the data

measured are non-numerical and the categories are
ranked. The data show that most students like school
quite a bit.

MHR  Data Management 12 Solutions 3

These are numerical, continuous data since the data are
numbers that can be any value. The data show that most
students have an SES Recode between 0.1 to 0.7.

These are numerical, discrete data since each value

can only be a distinct whole number. The data
indicate that most students have a general self-score
of 15.

These are categorical, nominal data since they are

non-numerical and the categories are not ranked.
These data suggest that most students intend to
complete a university degree.

The only data that cannot be represented as a graph

is the child ID number. There is simply one per
student.

4 MHR  Data Management 12 Solutions

Chapter 5 Section 1 Example 3 Your Turn Page 202

Answers may vary.

a) For #vacation, tweets go up and down. In the first 24-h period, tweets are up at 12:00 pm,
6:00 pm, 4:00 am, and 11:00 am.
For #math, tweets go up and down. In the first 24-h period, tweets are up at 3:00 pm, 1:00 am,
and 8:00 am.
For #Junos, tweets are almost non-existent until 4:00 pm. Then, there is a peak at 8:00 pm,
dropping to nothing again by 12:00 am.

b) Data seems to indicate that the Twitter topics #math and #vacation are not that common (low
constant volume) and only peak when there is an interest.

c) 8:00 pm, because that hashtag was tweeted less often from then on.

d) No, these graphs could not be compared as they are because there were so many more tweets
in the third graph. Even at 12:00 p.m., when the number of tweets for #Junos was near its lowest,
there were more #Junos tweets than there were at any time for #math.

Chapter 5 Section 1 R1 Page 203

Answers may vary. This suggests that predictions made by political experts are not reliable.
About half the time they will be correct and the other half they are incorrect.

Chapter 5 Section 1 R2 Page 203

Answers may vary. Positive: Availability, ease of backup, lower risk of loss, and not susceptible
to viruses. Negative: Access speed, requires Internet connection, environmental impact with
electricity use and carbon emissions of data centres.

Chapter 5 Section 1 Question 1 Page 203

Answers may vary.

a) While both continuous and discrete data can be numeric, continuous data can have any value
within a given range but discrete data can have only certain separate values. For example, heights
or students is continuous but the number enrolled in each class is discrete.

b) While both ordinal and nominal data are qualitative, ordinal data can be ranked but nominal
data cannot. For example, a rating scale for a cookie taste test compared to type of cookie
preferred.

c) Numerical data is quantitative and can be continuous or discrete, while categorical data is
qualitative. For example, how many movies you see in a year versus the types of movies you like.

Chapter 5 Section 1 Question 2 Page 203

a) The table contains continuous numerical data in the Armspan column and in the Height
column. The remaining table columns contain categorical nominal data.

MHR  Data Management 12 Solutions 5

b) The table contains both types of numerical data: continuous data in the Reaction Time and
Travel Time columns and discrete data in the Number of Languages Spoken and Number of
Siblings columns. The remaining table columns contain categorical nominal data.

Chapter 5 Section 1 Question 3 Page 204

Chapter 5 Section 1 Question 4 Page 204

a) Months are denoted by number: Jan =1, Feb = 2, and so on.

b) The general trend shows an increase in pageviews over the first few months

c) A possible reason for the dip in pageviews in the months of June and July is that school was
not in session.

d) Answers may vary. I predict the number of pageviews expected in December will be about
1100.

Chapter 5 Section 1 Question 5 Page 204

a) There are different scales on the two vertical axes (one for Canada, one for the U.S.), but they
both show that auto sales are nearing pre-crisis levels. If there was only one scale used, Canada’s
auto sales would look like a relatively flat horizontal line.

b) Answers may vary. This graph does not show the dramatic drop in sales as well.

6 MHR  Data Management 12 Solutions

Chapter 5 Section 1 Question 6 Page 204

Answers may vary.

a) Seasonally adjusted data reveals underlying trends in data that seasonal movements tend to
mask.

b) I think that sales are compared to the same month in the previous year for the same reason
why seasonally adjusted data is used to reveal an underlying trend.

Chapter 5 Section 1 Question 7 Page 205

Answers may vary.

a) While superhero views are high, the number of hours watched barely register in the bar
graph.

b) Rewrite the Views column data in a different unit, say millions, so that they have more
meaning

c) I chose Thor, Captain America, and Iron Man as fictional characters.

Thor: Interest over time has the greatest spike in November 2013. Regional interest puts Iceland
as the country with the greatest interest.
Captain America: Interest over time has the greatest spike in July 2011. Regional interest puts
Brunei as the country with the greatest interest.
Iron Man: Interest over time has the greatest spike in May 2013 with a close second in May 2010.
Regional interest puts the Philippines as the country with the greatest interest.

MHR  Data Management 12 Solutions 7

Chapter 5 Section 1 Question 8 Page 205

Answers may vary.

a) I entered Jackie as a girl’s name. The US popularity of my name peaked in the 1960s with
671 per million babies.

b) I would need the number of students at my school with that name and the total number of
students to determine a number per million for comparison.

c) Roy peaked in the 1890s. Marilyn and Joan both peaked in the 1930s.

d) Madison, Jackson, and Avery

Chapter 5 Section 1 Question 9 Page 205

a) Over time, School A appears to be doing the best, as the results steadily increased each year.

b) Answers may vary. School A Positive: 84% of students maintained standard. School A
Negative: 5% of students dropped from standard.
School B Positive: 26% of students rose to standard. School B Negative: 19% of students never
met standard.
School C Positive: 80% of students maintained standard. School C Negative: 13% of students
dropped from standard.
Based on the cohort data, School B appears to have the best results. This school has the greatest
percent that rose to standard.

Chapter 5 Section 2 Principles of Data Collection

Chapter 5 Section 2 Example Your Turn Page 209

a) Since you survey anyone who is on the street, the data are collected in a way that is easily
accessible. So, this is a convenience sample. The entire population is not represented since only
people in that area at that time are surveyed.

b) Since 200 students in the same proportions as the enrollment in each department are selected,
this is a stratified sample. Everyone has an equal opportunity to be selected since the students are
randomly selected.

c) This is a cluster sample because the population is swim clubs in Ontario, a number of clubs
are randomly chosen, and every member in those clubs are surveyed. Since there is random
selection in the first set of groups and sections, all groups and sections are equally likely to be
chosen.

d) This is a simple random sample. The draw includes all possible basketball players, so each
person has an equally likely chance of being selected.

e) Since the names are in alphabetical order, and every ninth person is chosen, this is a
systematic sample. As long as the first person is chosen randomly, everyone has an equal chance
of being selected.

8 MHR  Data Management 12 Solutions

f) Since the students choose whether or not they provide ideas for activities, it is a voluntary
response survey. Even though all students are invited to respond, the data will reflect responses
only from those who are interested. Thus, the entire population will not be represented in the
sample.

g) This city’s population has a hierarchy of organization and random sampling occurs at each
level of the organization, so this is multistage sampling. Each household is equally likely to be
chosen since there is random sampling at every level.

Chapter 5 Section 2 R1 Page 210

Answers may vary. Both a cluster sample and a multistage sample divide the population into
groups, just differently. Next in a cluster sample, a random number of groups are chosen and each
member of the chosen groups is sampled. A multistage sample takes the group subdividing
further by creating a hierarchy and choosing a random sample at each level.

Chapter 5 Section 2 R2 Page 210

Answers may vary. If a population is biased one way or another, that will be evident in the
smaller sample size.

Chapter 5 Section 2 Question 1 Page 210

a) Since the belts on the factory line are ordered, the selection time in each hour is random, and
the next 10 belts are chosen, this is a systematic sample.

b) This is a simple random sample. Each person in the database has an equally likely chance of
being selected.

c) Since members choose whether or not to complete the survey, it is a voluntary response
survey.

d) Since 75 females, the same proportion as the female golf course members, are selected, this is
a stratified sample.

Chapter 5 Section 2 Question 2 Page 210

A population is all the individuals in a group being studied, while a sample is a group selected
from the population. For example, “all students in a school” is the population and a sample is
“one class in the school.”

Chapter 5 Section 2 Question 3 Page 210

Answers may vary.

#SimpleRandomSample each member of population has equal chance of being selected.
#SystematicSample orders population and people are chosen at regular intervals.
#StratifiedSample has same proportion of members from each group as population does.
#ClusterSample divides population into groups, randomly selects groups, and samples each
member of chosen groups.
#MultistageSample divides population into a hierarchy and chooses random sample at each level.
#ConvenienceSample choose members of population who are easy to access.

MHR  Data Management 12 Solutions 9

#VoluntarySample people choose whether to participate so contains members heavily for or
against topic.

Chapter 5 Section 2 Question 4 Page 210

Answers may vary. Three sample methods that could be used to conduct a school survey are
stratified, cluster, or voluntary. Stratified could be based on grade level representation or gender
representation from the given data. Cluster could be based on random class sample at each grade
level. Voluntary allows students to choose whether to respond.

Chapter 5 Section 2 Question 5 Page 210

Answers may vary. I would conduct systematic sampling by randomly selecting a time each hour
and then testing the next 10 batches of candy mix on the factory line.

Chapter 5 Section 2 Question 6 Page 211

Answers may vary. Audrey Tobias is an 89-year-old peace activist who refused to fill out the
2011 census because of its link to a U.S. military contractor. She faced a criminal charge under
the Statistics Act, which makes refusing or neglecting to fill in the census form an offence
punishable by a $500 fine and up to three months in prison. Her lawyer argued forcing her to
complete the census that is processed using software from Lockheed would violate her freedoms
of conscience and free expression. She was acquitted due to reasonable doubt as to her intent at
the time of refusal.

Chapter 5 Section 2 Question 7 Page 211

a) These are both random sampling techniques that divide the population into groups, but
differently. Stratified uses same proportions as groups in population, whereas multistage uses
sequential levels of sampling by subdividing the group into smaller groups. For example,
stratified sampling to survey teachers may use gender groups for representation. On the other
hand, multistage sampling could first select from a group of high schools, then departments, then
teachers.

b) These are both non-random sampling techniques that can produce unreliable results because
people who respond are not representative of the population. While convenience sampling bases
selection on easy access, voluntary sampling is comprised of only those who choose to
participate. For example, convenience sampling to survey best movie could poll people at a
movie theatre. On the other hand, voluntary sampling could ask people to vote online.

Chapter 5 Section 2 Question 8 Page 211

a) The population is customers of the car dealership, of which there are 3757.

b) Answers may vary. First, determine the breakdown of customers by type of vehicle
purchased. Then, calculate the number of calls to make for each type of vehicle. Rounding is an
issue, resulting in 251 calls. I decided to round the economy car number down.

10 MHR  Data Management 12 Solutions

 Number of Percent of Number of
Type of Vehicle Customers Customers Calls to Make
SUV/truck 858 22.8% 57
Minivan 1213 32.3% 81
Midsize car 478 12.7% 32
Economy car 987 26.3% 65
Sports car 221 5.9% 15

c) Answers may vary. Perhaps they hope to get more responses than simply those that may
volunteer. Perhaps this is more cost effective.

d) Answers may vary. Ensure actual customer is surveyed not just a person living at that
household.

Chapter 5 Section 2 Question 9 Page 211

a) For 92% acceptable load, it is worth 0.92(41.9915)($94.40), or about $3646.88.

b) Answers may vary. If all 150 loads have 67% tomatoes that are acceptable, the farmers
minimum income would be 150(0.67)(41.9915)($94.40), or about $398 381.76. If all 150 loads
have 92% tomatoes that are acceptable, the farmers maximum income would be
150(0.92)(41.9915)($94.40), or about $547 031.67.
So, the farmer’s income from tomatoes can vary by almost $150 000 in any given year. Farmers
should save some of the income from good years so that they have money in case the next year is
not as successful.

c) The sampling method used is a simple random. Answers may vary. I do not think that one
random sample for the year is accurate enough to be reliable. Perhaps random samples should be
taken from every tenth load.

Chapter 5 Section 2 Question 10 Page 211

Answers may vary. Positive: Canadians can no longer be forced to divulge detailed personal
information under threat of prosecution.
Negative: Since the National Household Survey is not mandatory, data collected on a voluntary
basis will not be as accurate as the mandatory long form, and there will most likely be fewer
respondents.

Chapter 5 Section 2 Question 11 Page 211

Aerial surveys are done to estimate the number of moose. The park is divided into fifty-seven 25-
km2 plots representing either high-quality or low-quality winter moose habitat. Twenty percent of
the 57 blocks are randomly selected and moose are counted to estimate the population. This
sampling technique is similar to cluster sampling. In this case, the habitat is divided into areas, a
number of areas are randomly chosen, and population is counted in those areas.

MHR  Data Management 12 Solutions 11

Chapter 5 Section 3 Collecting Data

Chapter 5 Section 3 Example 1 Your Turn 1 Page 214

a) This is an observational study because there are no controls for the number of colds and the
subjects are not randomized.

b) Answers may vary. A control group that does not go to the gym is needed, then the treatment
group consists of those who do go to the gym. The effect being studied is whether going to the
gym reduces the number of colds one gets.

Chapter 5 Section 3 Example 1 Your Turn 2 Page 214

a) The control group contains plants growing in water with neutral pH. The experimental groups
are those with plants grown using water with increasing acidic pH levels.

b) Answers may vary. I think groups of plants were used to help eliminate any previously
diseased plants.

Chapter 5 Section 3 Example 2 Your Turn Page 216

Answers may vary.

a) The question includes only three options. Instead, make this an open question and allow
respondents to fill in the blanks.
What is your favourite game system?

b) This question is biased. It suggests that leadership training is necessary to run a business.
Rephrase it to make it neutral.
Do you think that leadership training would help you run a business? ■Yes ■ No ■ No opinion

c) This question asks for an opinion about two things. Either ask two separate questions, or give
options that cover all the situations.
How important do you think speed of service is?
■Very important ■Important ■Somewhat important ■Of little importance ■ Not important
How important do you think quality of service is?
■Very important ■Important ■Modestly important ■Of little importance ■ Not important

Chapter 5 Section 3 Example 3 Your Turn Page 218

Answers may vary. I created a survey to compare how much males and females like soccer.

12 MHR  Data Management 12 Solutions

Chapter 5 Section 3 R1 Page 219

Answers may vary. To get accurate and unbiased information, it is important to ask clear and
concise survey questions.

Chapter 5 Section 3 R2 Page 219

Answers may vary. I think a placebo group is also studied to account for effects from treatment
that do not depend on the treatment itself, such as knowledge and expectations of treatment. A
placebo group gives researchers a comparison group to determine whether the treatment itself had
any effect.

Chapter 5 Section 3 R3 Page 219

Answers may vary. Some ethical issues are informed consent, voluntary participation, avoidance
of harm, and participant confidentiality and data protection.
Informed consent must be acquired at several points in time.
Voluntary participation must consider the person’s well-being when making the decision.
Avoidance of harm can lead participants to change their behaviour due to influence of feedback.
Participant confidentiality and data protection must be adhered to since observations of change
across time can generate scenarios that would provide clues to an individual’s identity.

Chapter 5 Section 3 Question 1 Page 219

The rating scale of usually true, occasionally true, and sometimes true is unacceptable. These are
all “positive” versions. Answer C.

Chapter 5 Section 3 Question 2 Page 219

Experimental studies must have variables controlled throughout the study. Answer B.

Chapter 5 Section 3 Question 3 Page 220

a) This question asks about two things. Ask two separate questions.
Do you exercise daily? ■Yes ■ No
Do you get enough sleep at night? ■Yes ■ No

MHR  Data Management 12 Solutions 13

b) The statement is biased and the question is unrelated. Rephrase it to make it neutral.
Do you think that the Star Wars saga is one of the best science fiction stories of all time?
■Strongly agree ■Agree ■Don’t know ■Disagree ■Strongly disagree

c) The question includes only three options. Instead, make this an open question and allow
respondents to fill in the blanks.
What is your favourite type of music?

d) This statement is confusing by the choice or wording. Make it concise.

How do you feel about the following statement: We should increase the number of recycling days
in the school?
■Strongly agree ■Agree ■Don’t know ■Disagree ■Strongly disagree

Chapter 5 Section 3 Question 4 Page 220

Answers may vary. Example: How often do you walk to school?

Chapter 5 Section 3 Question 5 Page 220

a) Since there are no controls and subjects are not randomized, this is an observational study.
The effect being studied is whether magnetic therapy reduces pain.

b) Since there are no controls and subjects are not randomized, this is an observational study.
The effect being studied is whether lying in bed for an extended period of time exhibits similar
issues to weightlessness in space.

c) Since there is randomization and some aspect of control, this is an experimental study. The
effect being studied is whether handedness affects math grades.

Chapter 5 Section 3 Question 6 Page 220

Answers may vary.

a) Approximately 410 people had zero computers in their home.

b) The portion highlighted in red represents those who also have not used the Internet.

c) I think that the 0 computer category will either not exist or be very small, and the 1 and 2
computer categories will greatly increase. Similarly, the no for connected to the Internet will
either disappear or be very small, and the yes will greatly increase. Also, all of the red portions
should be much smaller, because most people have used the Internet.

d) The graphs show the number that used email in the last 12 months, word processed in the last
12 months, and used a computer in their main job in the last 12 months with those that responded
no when asked whether they had used the Internet highlighted in red.

14 MHR  Data Management 12 Solutions

Chapter 5 Section 3 Question 7 Page 221

Answers may vary.

Which is your favourite sport?
■Tennis ■Hockey ■Basketball
How do you like K-pop music?
■Pretty Good ■Good ■Great ■Fantastic ■Awesome
How important do you think exercise and eating healthy are?
■Very important ■Important ■Modestly important ■Of little importance ■Unimportant

MHR  Data Management 12 Solutions 15

Chapter 5 Section 3 Question 8 Page 221

Answers may vary.

a) It explains that the data contain 43 attributes from Statistic Canada’s 2001 Census of
Population for Canada. It also provides a list of the attributes with a description of the meaning of
each.

b) The first graph shows all respondents were from Ontario. The second graph shows that the
greatest number of respondents were from Toronto.

c) The first graph shows that most people are of CBFP origins. The second graph shows that
most people’s mother tongue is English. The third graph shows that the place of birth for most
people is Ontario.

16 MHR  Data Management 12 Solutions

MHR  Data Management 12 Solutions 17
d) It may be important to have data on location and ethnicity to study cultural trends, health care
needs, and other social-economic issues.

Chapter 5 Section 3 Question 9 Page 221

Answers may vary.

How concerned are you about safety in sports?
■No opinion ■Not concerned ■A little concerned ■Concerned ■ Very concerned
What is your favourite classification of sports?
■Endurance sports ■Team sports ■Target sports ■Strength sports ■Athletics ■Other
How many hours a week do you watch sports?
■0–5 ■6–10 ■11–15 ■16–20 ■more than 20
What sport do you play?
What is your level of expertise in your sport on a scale of 1 to 5 with 1 being poor and 5 being an
expert?

18 MHR  Data Management 12 Solutions

Chapter 5 Section 3 Question 10 Page 221

Answers may vary.

Do you agree that cell phones should be silenced in movie theatres?
■Strongly agree ■Agree ■Undecided ■Disagree ■Strongly disagree
How often do you see a person using a cell phone while he or she is driving?
■Very frequently ■Frequently ■Occasionally ■Rarely ■Never
I find my mathematics class interesting.
■Almost always true ■Usually true ■Occasionally true ■Sometimes true ■Never true
How well do you like the latest sci-fi movie release?
■Like a lot ■Like a little ■Do not like or dislike ■Dislike a little ■Dislike a lot
How important do you think global warming is?
■Very important ■Important ■Modestly important ■Of little importance ■Unimportant

Chapter 5 Section 3 Question 11 Page 221

Answers may vary.

Chapter 5 Section 3 Question 12 Page 221

a) They likely conducted an observational study.

b) Answers may vary. A control crop (group) that uses chemical fertilizers is needed, then the
treatment crop (group) uses natural fertilizers. The effect being studied is whether natural organic
crops yield less than traditional crops.

Chapter 5 Section 3 Question 13 Page 221

a) Answers may vary.

How often do you eat breakfast?
■Never ■Rarely ■Sometimes ■Most of the time ■Always
How many times a day do you eat fruits and vegetables?
■Less than 1 ■1 to less than 3 ■3 to less than 5 ■5 or more
How many sodas or pops did you drink yesterday?
■0 ■1 ■2 ■3 ■4 or more
What were your grades like last term?
■Mostly As ■Mostly Bs ■Mostly Cs ■Mostly Ds ■Mostly Fs
During the last term, how many days of school did you miss due to illness?
■0 ■1 ■2 ■3 ■4–5 ■More than 5

b) to d) Answers may vary.

Chapter 5 Section 3 Question 14 Page 221

Answers may vary. The subjects cannot be randomly assigned to a treatment group or control
group. It is not feasible to assign the “treatment” of academic success.

Chapter 5 Section 3 Question 15 Page 221

a) Answers may vary. Studies may seem questionable if the source does not seem reliable, if the
sample seems skewed, or if the methods seem biased. Sample topics:

MHR  Data Management 12 Solutions 19

Coffee can help people feel less tired and increase energy levels.
Several studies show that caffeine can boost the metabolic rate by 3-11%.
Coffee may lower your risk of type II diabetes.

b) “Investigation of the Effects of Coffee on Alertness and Performance during the Day and
Night”: This study was experimental. It examined the effects of coffee on performance and
alertness in the day and at night.
“Normal caffeine consumption: influence of thermogenesis and daily energy expenditure in lean
and postobese human volunteers”: This study was experimental. This measured energy
expenditure in a room respirometer.
“Coffee, Decaffeinated Coffee, and Tea Consumption in Relation to Incident Type 2 Diabetes
Mellitus A Systematic Review With Meta-analysis”: This study was observational. This was a
meta-analyses for the association between coffee, decaffeinated coffee, and tea consumption with
risk of diabetes.

c) This is an observational study. In it a positive association between coffee consumption and

all-cause mortality was observed in men and in men and women younger than 55 years. This
group should avoid drinking greater than four cups of coffee a day.
Answers may vary. The study population included a range of ages all of which completed a
medical history and lifestyle questionnaire, were given a baseline examination for mortality, and
present findings for all-cause mortality are consistent with those of earlier studies. However, the
results of recent studies have been highly variable. The non-coffee-drinking group may have had
a higher mortality risk not related to the consumption of coffee.

Chapter 5 Section 3 Question 16 Page 221

Answers may vary.

Entitled Meta-analysis of randomized controlled trials of statins versus placebo in patients with
heart failure, this study found that compared with placebo, statin therapy resulted in a statistically
significant reduction in the incidence of hospitalisation for worsening heart failure.

Entitled Initial Severity and Antidepressant Benefits: A Meta-Analysis of Data Submitted to the
Food and Drug Administration, this study examined all available data on antidepressant drugs
(fluoxetine, venlafaxine, nefazodone, and paroxetine) given to the FDA, including results from
clinical trials that the manufacturers chose not to publish at the time. The conclusion was that
patients on placebo improved just as much as those on the drugs. In other words, that the overall
effect of these antidepressants was below the recommended criteria for clinical significance

Chapter 5 Section 4 Interpreting and Analysing Data

Chapter 5 Section 4 Example 1 Your Turn Page 225

Answers may vary.

a) It appears that the average domestic fare decreased from 2008 to 2009 by around $30.

b) You can compare average domestic fares for 2008 and 2011 by individual city, by province,
by cities within a province, or for Canada.

20 MHR  Data Management 12 Solutions

Chapter 5 Section 4 Example 2 Your Turn Page 227

a) For each song, you are given its name, the artist, the composer, the album it came from, the
genre, the size of the file, the length of the song, the year it was released, the sampling bit rate,
and the number of times it was played.

b) Answers may vary. Assuming that time is in seconds, the most common length for a song is
from 210 s to less than 230 s.

c) No. Select each bar in the genre graph and observe the distribution in the bit rate graph.
Pop is mostly sampled at 320 bits per second. The two main sampling rates for rock are 128 and
192 bits per second. Alternative is mostly sampled at 128 bits per second. Hip hop is mostly
sampled at 192 bits per second. 80s music is typically sampled at 160 or 192 bits per second.
Electronic is split between 160 and 320 bits per second, while R&B is split between 160 and 192
bits per second. Finally, metal is only sampled at 320 bits per second.

d) Answers may vary. If you know how long the song is and the sampling rate, then you
determine the size of the file.

Chapter 5 Section 3 Example 3 Your Turn Page 229

a) This is an experiment because the student is controlling the drop height of the ball.

b) One condition is controlled, the drop height of the ball, and results, bounce height, are
recorded.

c) Bounce height appears to be on average about 70% of the drop height.

d) Use a graphing calculator. The relationship appears to be linear.

e) Yes, the data can be used to predict the bounce height for a drop height of 130 cm by using a
line of best fit.

Chapter 5 Section 4 R1 Page 230

Websites for the NHL, MLB, NBA, and NFL are like databases because each provides an
organized store of records containing information on players, teams, and games.

Chapter 5 Section 3 R2 Page 230

Answers may vary. I believe that secondary data is more visible to the public because we
typically are exposed to someone else’s use of primary data.

MHR  Data Management 12 Solutions 21

Chapter 5 Section 4 Question 1 Page 230

The height of each person in your class is microdata. Answer B.

Chapter 5 Section 4 Question 2 Page 230

Table 427-0004 contains information about vacations. Answer C.

Chapter 5 Section 4 Question 3 Page 230

The topic of Table 128-0014 is Electricity generated from fossil fuels

The topic of Table 203-0026 is Survey of household spending (SHS), household spending, by age
of reference person.
The topic of Table 427-0004 is Number of international tourists entering or returning to Canada,
by province of entry.
The topic of Table 361-0013 is Spectator sports, event promoters, artists and related industries,
summary statistics, by North American Industry Classification System (NAICS)

Chapter 5 Section 4 Question 4 Page 230

a) Since the town collects data directly from the households, this is primary data.

b) Since the town publishes the average number of people per house on each street, the data has
been summarized and is secondary data.

c) If you assume that the newspaper collected the information and then reports the salary of
every city council member, not manipulated, this is primary data. If this is a matter of public
record and they are simply reprinting it, it is secondary data.

d) Since the magazine lists the total number of home sales each month for the last year, the data
has been summarized and is secondary data.

e) Since the table shows actual microdata, this is primary data.

f) Since the table shows summarized data, it is secondary data.

Chapter 5 Section 4 Question 5 Page 231

Answers based on 2011 data.

a) Answers may vary. In Kingston, Ontario, about 1% of people are in my age group of 17
years.

b) Answers may vary. In Kingston, Ontario, the largest age group in my city is 20 to 24 years.

c) In Toronto, about 1% of people are in my age group of 17 years. The largest age group in
Toronto is 25 to 29 years.

d) Answers may vary. For Kingston: population was 123 363 people; population density per
square kilometre was 273.4; the percent of people married or living with a common-law partner
was about 54%.

22 MHR  Data Management 12 Solutions

For Toronto: population was 2 615 060 people; population density per square kilometre was
4194.5; the percent of people married or living with a common-law partner was about 51%.
Toronto has a much larger population, and the population density is more than twice that of
Kingston. However, the percent of people married or living with a common-law partner is a bit
less.

Chapter 5 Section 4 Question 6 Page 231

10
a) The table contains data for 10 earthquakes over magnitude 4. So, , or about 21% of
48
earthquakes over magnitude 4 occurred in North or South America.

b) The only earthquake above magnitude 4 in Canada occurred in the Queen Charlotte Islands.
This earthquake was the largest recorded in North or South America in 2012.

c) I chose a dot plot to compare magnitudes.

Collection 1 Dot Plot

4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0

Magnitude

d) This graph shows the location of the earthquakes.

Answers may vary. I think this is called a heat map because of the use of colour to denote severity
of magnitude.

MHR  Data Management 12 Solutions 23

e) There is a concentration of large magnitude earthquakes in South America.

f) There were more large magnitude earthquakes in 2013.

Chapter 5 Section 4 Question 7 Page 231

a) The information stored about each image is the date, time, exposure time, F-stop, Exp
program, metering mode, whether flash was used, focal length, ISO speed, orientation,
dimensions, total pixels, and file size in kilobytes.

24 MHR  Data Management 12 Solutions

b) Answers may vary.
Images Dot Plot

Images Dot Plot

0 400 800 1200 1600 0 2 4 6 8 10 12 14 16

ISO_Speed F_Stop

Images Dot Plot

0 2 4 6 8 10 12 14 16 18
Total_Pixels (millions)

MHR  Data Management 12 Solutions 25

This information tells the photographer how often a certain speed film or f-stop is used, the most
common image sizes, how often the flash is fired, the most common metering mode and
orientation.

c)

The camera can use the number of pixels to determine the corresponding file size. Then, divide
the remaining storage capacity by the file size and estimate how many more photos can be taken
at that setting.

Chapter 5 Section 4 Question 8 Page 232

a) The numbers in the table are primary data from weather stations. The table represents
secondary data for anyone else who uses it.

b) While Stratford, Ont gets 1064 mm of rain every year, Hartley Bay, BC gets over four times
more rain at 4549 mm per year and Henderson Lack, BC gets over eight times more rain at 9082
mm per year.

c) With locations organized by province, it is easy to see that the six highest rainfall locations
are all in BC.

26 MHR  Data Management 12 Solutions

Chapter 5 Section 4 Question 9 Page 232

a) Since this is a collection of data from various scientists, this is secondary data.

b) These data are the result of a study. This is collected data about a natural phenomenon.

c) Answers may vary. Data on number of impacts per region may help scientist predict the
location on the next impact.
It could be simply because North America is the third largest continent.

d) Europe has the lowest portion of craters drilled. This is most likely because they are at
inaccessible locations.

e) Answers may vary. Ordering the Age_Ma column of the table, it is not obvious that the older
the impact, the larger the diameter. The graph confirms this.

MHR  Data Management 12 Solutions 27

Chapter 5 Section 4 Question 10 Page 232

Answers may vary.

a) Once the Play button is clicked, the various sizes and colours of dots begin to change size and
location. This shows that both income per person and life expectancy increased over time.

b) The size of each circle represents the population of the country and the colour represents its
geographic region. The size is in relation to the world population.

c) Each country’s story is told in regards to income per person and life expectancy over time.
Canada for the most part has continuously increased in both income per person and life
expectancy.

d) I looked at child mortality rates from 1800 to 2012. In general, it shows that child mortality
rates decreased with increase in income per person over time.

Chapter 5 Section 5 Bias

Chapter 5 Section 5 Example 1 Your Turn Page 234

a) This is response bias. Because you are asked how many words per minute you can read, it is
likely you will boost that number to give a good impression.

b) Since you are limiting whom you ask to parents of school-age children, it is likely the
respondents will all be in favour of safety lanes. This is sampling bias.

c) Usually, the people who respond to a voluntary survey are those with extreme opinions.
People who are indifferent or do not think the topic is important may not respond to the text. This
is non-response bias.

d) This is measurement bias, because it is a leading question.

Chapter 5 Section 5 Example 2 Your Turn Page 236

Answers may vary.

a) This graph is misleading is several ways. There is no vertical scale and the points suggest a
linear relationship. However, the number of jobs lost does not change be a constant amount. Also,
the horizontal scale is not consistent. Increments vary from nine months to almost ten years.

b) While the rate may have dropped from July to September, it did not drop in October. This
graph is further misleading by not having the vertical scale start at 0.

c) This information is most likely anecdotal and cannot be verified. In addition, no actual data
on the amount of weight lost is given or whether consumers were on a diet and exercise regimen

Chapter 5 Section 5 Example 3 Your Turn Page 237

Answers may vary.

a) First, determine the amount spent per household. For 2008-2012, there were 115 226 802
households in the United States, and in 2011 there were 13 320 614 households in Canada.

28 MHR  Data Management 12 Solutions

 8000 000 000
 62.2338...
115 226 802  13 320 614
This works out to about $62.23 per household. For countries that do not celebrate Halloween, this
is a lot to spend on candy, costumes, and decorations. However, for Canadians and Americans
this is not a lot of money.

b) Using data from the Internet, I found Apple’s gross profit in 2010 was $26.71 billion and in
2011 it was $45.63 billion. So, $19 billion is a large increase of about 59%.

Chapter 5 Section 5 Example 4 Your Turn Page 239

a) The Canadian teams are distinguished from the others by having a white background and the
logo is inserted inside the bar.

b) The team with the fewest North Americans would have the smallest percent of Canadians and
Americans.

c) Answers may vary. You could argue that hockey is Canada’s pastime and not that of the
United States based on the number of Canadian-born players in the NHL and on each team.

d) Answers may vary. It is difficult to determine the actual percent of American-born players on
each team.

e) No. This infographic is not free of bias. Because of presentation and information given, it is
definitely slanted towards Canadian-born players.

f) Answers may vary. Canadians rule in the NHL.

Chapter 5 Section 4 R1 Page 240

Answers may vary. Yes. It possible to make misleading claims even if the data are free from bias
by displaying it in a distorted fashion.

Chapter 5 Section 4 R2 Page 240

Answers may vary. You can “lie” with statistics simply by omitting other information. It is also
possible the statistics are based on faulty experiments or studies.

Chapter 5 Section 4 R3 Page 240

Answers may vary. The company is claiming that the consumer is at fault. Just because the
product is called vitaminwater ® this in no way should imply that it contains vitamins and
consumers must read the fine print labels.

Chapter 5 Section 5 Question 1 Page 240

The question, “How important is health care compared to other social issues?” is not a leading
question. Answer D.

MHR  Data Management 12 Solutions 29

Chapter 5 Section 5 Question 2 Page 240

Answers may vary.

a) This is response bias. Students may feel pressured to raise their hand even if they did not
understand the lesson. The teacher could make the response anonymous by asking students to
write their answer on a piece a paper, all of which are collected in a hat.

b) Usually, the people who attend city council meeting have extreme opinions. In addition, a
large proportion of citizens may not be able to attend. This is non-response bias and sampling
bias. Perhaps several open public meetings would gather a more accurate opinion.

c) Since only followers of the news program will likely respond to the poll, it is sampling bias
since the respondents likely would not properly represent the population. To better represent the
population, the poll needs to be available in other ways.

d) This is response bias. Because you are asked how much you like to exercise, it is likely you
will say more than the truth to give a good impression. Instead ask how often the person
exercises.

e) Since you are limiting whom you ask to customers of the hunting equipment store, it is likely
the respondents will overwhelmingly want to lower the cost of registering a gun. This is sampling
bias. To better represent the population, the petition needs to be available in other ways.

Chapter 5 Section 5 Question 3 Page 240

Answers may vary. I chose measurement bias.

A survey question asks, “Who is the best supercross rider of all time, Ricky Carmichael or James
Stewart?

Chapter 5 Section 5 Question 4 Page 241

Answers may vary.

a) The statement does not explain what the typical number of headaches was, nor does it
identify other potential factors (diet, water consumption, weather).

b) Because of the angle of this 3D representation, the two sections that indicate people who like
math look significantly larger than the two sections that indicate people who do not like math.

c) The break in the vertical scale exaggerates the change in population. It appears to have more
than doubled looking at the height of the bar, but it actually only changed by 3 million.

d) It is very difficult to determine actual values from the use of icons.

e) The break in the vertical scale exaggerates the change in units sold, and the horizontal scale
does not show consistent monthly increments.

Chapter 5 Section 5 Question 5 Page 241

a) Increasing polling booths in strong support areas and decreasing them in low support areas
will result in over-representation of supporters on the party in power. This is measurement bias.

30 MHR  Data Management 12 Solutions

b) Answers may vary. Use percentage points to make the win sound bigger. For example, the
Conservative Party wins over the NDP by 9 percentage points.

c) Voter turnout for the 2011 election was 61.1%. The population eligible to vote was 24 257
592.
Percent of Voting
Political Party Population That Voted
Conservative Party 24.2
NDP 18.7
Liberal Party 11.5
Bloc Québécois 3.7
Green Party 2.4

This leaves 38.9% or the voting population that did not vote, which is greater than any percent
that voted for a party. This could be non-response bias in that people chose not to vote because of
indifference or other reason.

d) Answers may vary. In 1898, not even half of the population voted. This might have resulted
in under representation of some groups and over representation of others.

e) Answers may vary.

Chapter 5 Section 5 Question 6 Page 242

a) Determine the pieces of mail per dwelling per delivery day. The 2011 Census counted 13 320
614 occupied private dwellings.
9 800 000 000
 2.8296...
13 320 614  (52  5)
That represents about three pieces of mail per dwelling per day, which seems typical. So, the
number 9.8 billion is not significant.

b) Determine the number of hours watched per viewer. There are more than 1 billion unique
user visits to YouTube each month. So, there are about 12 billion users per year.
70 000 000 000
 5.8333...
12 000 000 000
Each viewer watches about 5.8 h of video each year does not seem like a lot. So, the number 70
billion is not significant.

c) Since Clayton Kershaw is the highest paid player in Major League history, and he earns more
than 15% more than the next highest paid player, this number is significant.

d) Compare this with last year’s national debt of $602.4 billion; it is an increase of about 8%.
This is a significant number.

Chapter 5 Section 5 Question 7 Page 242

a) Answers may vary. The sun-like design and bright colours make is appealing.

MHR  Data Management 12 Solutions 31

b) Canada ranks 29th. Answers may vary. This could be because of environmental conditions,
lack of federal spending, public attitudes, and other cheaper established forms of generating
electricity.

c) The data visually connect to the topic of solar power by the sun-like shape and use of “sun”
colours.

d) Answers may vary. I believe the editors are trying to imply that Canada is far behind in its
use of solar energy and needs to improve its standing.

Chapter 5 Section 5 Question 8 Page 242

Answers may vary. Shocking: Cervical cancer rates up by 25%.

Neutral: Cervical cancer rate changes from 1.6% in 2013 to 2% in 2014.

Chapter 5 Section 5 Question 9 Page 242

Answers may vary. The push poll is a form of telemarketing-based propaganda disguised as a poll
that attempts to influence or alter the view of voters. It is viewed as a form of negative
campaigning.

Chapter 5 Section 5 Question 10 Page 242

Answers may vary. The Hawthorne effect refers to the fact that people will modify their behavior
simply because they are being observed. A simple example would be a typically slow worker that
works faster and does a better job when his or her boss or manager is watching him or her.

Chapter 5 Section 5 Question 11 Page 242

Answers may vary.

a) Confirmation bias is a tendency for people to favour information that confirms their
preconceptions regardless of whether the information is true or not. For example, watching
certain TV news programs or only visiting websites that express your opinion.

b) Observational selection bias is that effect of suddenly noticing things you did not notice that
much before and you wrongly assume that the frequency has increased. For example, when one
of your car’s headlights is out and suddenly you noticed other cars with the same problem.

Chapter 5 Section 5 Question 12 Page 242

Answers may vary.

a) This is measurement bias. This is a leading question; the wording of the question can affect
the outcome by influencing someone’s answer. Change the question to simply, “How important is
it for you to use high-tech farming methods?”

b) The firm presents the information in this misleading manner to exaggerate their growth and to
try and convince potential new clients to “get onboard.”

Chapter 5 Section 5 Question 13 Page 243

Answers may vary.

32 MHR  Data Management 12 Solutions

a) I was surprised to see that the majority of movies are sci-fi/fantasy. Also, the highest ranked
movie has made over $100 million more than the movie ranked at number two.

b) The greatest number of movies was released after 2010. These movies have an unfair
advantage over much older movies, which had cheaper ticket prices.

c) This graph is almost the reverse of the other one.

d) Answers may vary. The top movie on the unadjusted list drops to number three on the
adjusted list. Also, the adjusted list contains quite a few different movies and only five movies
were released in the 2000s.

e) Answers may vary.

Chapter 5 Section 5 Question 14 Page 243

Answers may vary.

a) There is an anti-vaccination movement with celebrity supporters who believe there is a link
between vaccines and autism.

b) The possible link between vaccines and autism is in part based on a very small and flawed
study published in the UK back in 1998. The study was retracted in 2004, and the CDC, the
Institute of Medicine of the National Academy of Sciences, and the UK National Health Service
have all concluded that there is no evidence of a link between the MMR vaccine and autism.

Chapter 5 Section 5 Question 15 Page 243

Answers may vary. A double blind study is one in which neither the subjects nor the scientists are
aware of who is receiving the medication and who is receiving a placebo. This helps to prevent
results from being influenced by the placebo effect or observer bias.

Chapter 5 Review

Chapter 5 Review Question 1 Page 244

Answers may vary.

a) numerical, discrete: number of students in a class

b) numerical, continuous: height of students in a class

c) categorical, ordinal: answers to a question that uses a rating scale

d) categorical, nominal: eye colour of students in a class

Chapter 5 Review Question 2 Page 244

a) I believe the peaks coincide with weekends, since they occur on a regular weekly basis.

b) Answers may vary. I estimate that they went over their limit and used 102 Gb.

MHR  Data Management 12 Solutions 33

c) Using the sum function in a spreadsheet, the family actually used 104.73 Gb.

d) The average daily Internet usage of the family is 3.491 Gb. Answers may vary. I do not think
this number is useful to the family. They almost always use less than this on weekdays and
significantly more on weekends.

Chapter 5 Review Question 3 Page 244

a) Since the user chooses whether or not they respond to the survey, it is a voluntary response
survey.

b) Since the researchers just picked three trees, the data are collected in a way that is easily
accessible. So, this is a convenience sample.

Chapter 5 Review Question 4 Page 245

a) There are two possible responses for age 15, 20, and 35.
How old are you?
■15 and below ■16–19 ■20–34 ■35–60 ■Above 60

b) This question is biased. It suggests that violence in video games could make people violent.
Rephrase it to make it neutral.
How do you feel about a violence rating system for video games?
■Strongly agree ■Agree ■Agree a little ■Don’t agree

c) Give a range of responses.

Do you like the new logo for the school?
■Like it a lot ■Like it ■Do not like or dislike ■Dislike a little ■Dislike a lot

Chapter 5 Review Question 5 Page 245

a) This is an experimental study. Researchers control the method of asking for donations and
make inferences based on those controls.

b) This is an observational study. The teacher is reviewing situations that have already occurred
and making an inference.

Chapter 5 Review Question 6 Page 245

a) The numbers in the table are primary data from whatever agency collected it. The table
represents secondary data for anyone else who uses it.

b) This is primary data since you are collecting it.

c) Assuming this is collected public information, it is secondary data.

Chapter 5 Review Question 7 Page 245

Answers may vary.

a) Any scenario where a small group that is not representative of the population is surveyed.

34 MHR  Data Management 12 Solutions

b) Any scenario where a limited list of answer choices is given or an opinion is stated before
asking the question.

c) Any scenario that influences an answer so as to avoid embarrassment or please the questioner.

d) Any scenario where a response is voluntary.

Chapter 5 Review Question 8 Page 245

a) The size of the various fruit icons is misleading. The rows are basically the same length, but
due to the size of the icons used each actually represent s a different number.

b) This 3-D pie chart is misleading because it is difficult to judge the sizes of each sector and no
percent data is given.

Chapter 5 Review Question 9 Page 245

Answers may vary,

Chapter 5 Test Yourself

Chapter 5 Test Yourself Question 1 Page 246

During one of these months, Facebook received approximately 870 million unique visitors.
1 1012
 1149.4252.....
8.7  108
This represents on average 1149 page views per visitor per month, or about 38 page views per
day. This is a significant number. Answer A.

Chapter 5 Test Yourself Question 2 Page 246

Systematic: Order data from highest to lowest, randomly select one item, and then choose every
5th value after that.
Put all the data into a bin and randomly select as many as you wish. This is not a valid sampling
method.
Stratified: Arrange the data into demographic groups, and then pick 100 from each group.
Voluntary: Send out a survey and ask people to fill it out and send it back.
Answer B.

Chapter 5 Test Yourself Question 3 Page 246

The first choice limits answer options.

The second choice is a good survey question.
The third choice is a leading question.
The fourth choice limits answer options.
Answer B.

MHR  Data Management 12 Solutions 35

Chapter 5 Test Yourself Question 4 Page 246

In the first choice, data represents averages, so it is aggregate data.

In the second choice, data represents about one class, so it is microdata.
In the third choice, data is summarized, so it is aggregate data.
In the fourth choice, data represents averages, so it is aggregate data.
Answer B.

Chapter 5 Test Yourself Question 5 Page 246

Answers may vary.

a) The sales representative will say whatever it takes to get you to buy a system. Each household
varies in is electricity usage. The representative cannot guarantee anything and should instead
offer an in depth assessment of the size of the system needed and projected payback timeline.

b) This scenario is response bias, because a student may respond falsely to avoid
embarrassment. The teacher could make the response anonymous by asking students to rate their
understanding on a scale of 1 to 5 on a piece a paper, all of which are collected in a bag.

Chapter 5 Test Yourself Question 6 Page 246

Answers may vary.

a)

b) The graph shows the amount of waste in kilograms per person per year by country. A person
in Denmark creates the most waste in a year, while a person in Finland creates the least.

36 MHR  Data Management 12 Solutions

Chapter 5 Test Yourself Question 7 Page 246

Answers may vary.

How often do you use a computer to access the Internet?
■ Once a week ■Every day ■Several times a day ■Other

How many hours a week do you spend online?

Where do you most frequently use a computer to access the Internet?

■School ■Library ■Home ■Friends home ■Other (please specify)

Why do you access the Internet? Check all that apply.

■Check email ■Use a search engine ■Play computer games ■Download music ■Social media
■ Other (please specify)

Chapter 5 Test Yourself Question 8 Page 247

Answers may vary.

a)

b)

MHR  Data Management 12 Solutions 37

Chapter 5 Test Yourself Question 9 Page 247

Answers may vary.

a)

b) I think that this is stratified sampling because of the number of people born in each province
approximately matches the percent of Canada’s population living in each province.
Birth Province Number of People Percent of Surveyed
NFLD 18 2.04%
PEI 10 1.14%
NS 36 4.09%
NB 16 1.82%
QUE 225 25.54%
ONT 301 34.17%
MAN 29 3.29%
SASK 33 3.75%
ALTA 80 9.08%
BC 133 15.10%

c) Assuming that the data was collected using stratified sampling, then the headline is incorrect.
The percent of Ontarians travelling is slightly higher than their relative percent of the population.
Actually, the percent of people from New Brunswick travelling is much greater relative to their
percent of the population.

38 MHR  Data Management 12 Solutions

 Home Province Number of Trips Percent of Surveyed
NFLD 3500 1.53%
PEI 1091 0.48%
NS 8115 3.54%
NB 5458 2.38%
QUE 60169 26.27%
ONT 90174 39.37%
MAN 7984 3.49%
SASK 8624 3.77%
ALTA 21558 9.41%
BC 22380 9.77%
Total 229053 100.00%

d) Provide the percent of the population that travels rather than just the number of trips.

Chapter 5 Test Yourself Question 10 Page 247

a) Answers may vary.

b) Answers may vary. A highly credible strategy should ensure that the sample reflects the
characteristics of the population, use a non-biased sampling technique (simple random,
systematic, stratified, cluster, or multistage), and use survey questions free of bias.
A not very credible strategy would use a sample that reflects only some portion of the population,
use a biased sampling technique (convenience or voluntary), and use survey questions that
contain bias.

MHR  Data Management 12 Solutions 39