Module Eight Lesson One Guided Notes Name_____________________

Module Eight Lesson One Introduction – PARAMETERS AND STATISTICS

Statistics is one of the branches of mathematics that is used all around you and you might not even realize it!
Have you ever seen a survey on Facebook? Did you answer the questions that were being asked? What were they
going to do with the data that was collected from the sample of people on Facebook that answered the
survey? Surveys completed by Facebook users can generate data that is used to change things from the color
scheme of a page to a campaign strategy!
In this section, you will explore the difference between a population and a sample and why samples are used to
represent the population. In addition, you will explore how the means and proportions from a sample can provide
important data about the population. Finally, you will learn why it’s important to communicate the results of a
study or experiment with the general population. Have fun!

LESSON OVERVIEW
You see statistics around you every day, you just might not realize it! In this section we are going to take a look at
why someone might ask you if you text and drive. Your answer would be part of a sample that allows researchers
to figure out the proportion of people that text and drive. What do they do with that research? Make new laws! In
this lesson you will be taking a look at the importance of gathering samples that produce reliable data from
surveys you are collecting or experiments that you are conducting.

Lesson Objectives & Standards

Learning Objectives—By the end of this lesson, you will:
1- Describe the difference between a population and a sample.
2- Explain why we use a sample to make inferences about a population.
3- Distinguish the difference and purpose between sample surveys, observational studies, and experiments.
4- Identify the parameter of interest in an experiment/scenario.
5- Given data the student will calculate sample statistic- mean or proportion.
6- Generalize the result of a statistical study to make conclusions about a population based on the sample.

KEY TERMS
Population - Everything or everyone being studied in an inference procedure.
Sample-   A subset of the population.  
Inferential Statistics/Inference - Using information from a sample to draw conclusions about corresponding
population.
Sample survey: A survey (or a questionnaire) that is given to a few members of the population. 
Observational studies- A method of collecting information about a population by observing subject without
imposing treatment (i.e., examining the health effects of smoking).
Experiments-A method of applying a treatment to a group and recording the effects.
Placebo - It is a “fake” treatment. 
Treatment Group - The group in an experiment that receives the treatment. 
Control Group - The group in the experiment that receives the placebo. 
Comparative experiment-   A control group is given a placebo to compare the reactions between the treatment
group(s) and the control group.
Statistical study- A study where data is collected and the statistics (means, proportions, etc.) are examined.
Parameter-Any numerical value that comes from data collected from a population. 
Statistic-Any numerical value that comes from data collected from a sample.   
Sample mean- The mean/average that is found from a sample.
Sample proportion -The proportion that is found from a sample.  
Population mean - The mean/average that is found from a defined population. 
Population proportion-The proportion that is found from a sample. 
Census - Taking a sample of every individual member of the population. For example, when the entire population
of the United States is counted that is a census.

2
Notes One - Population Vs. Sample

Population is – everything or everyone being studied in an inference procedure.

Key here is that you can define the population
1) you want to define the population you are going to study.
Examples of a defined population:
 Everyone at your high school
 Everyone that lives in Raleigh
 All girls in 10th grade
 All of the soccer players in North Carolina
Census is -taking a sample of every individual member of the population. For example, when the
entire population of the United States is counted that is a census.
If your high school was your population what would the census be?
 The census would be everyone at your high school
 This is often hard to gather, because it might be hard to reach everyone at your school!
Thought question: Why do you think a census is impractical?
Sample is - a subset of the population
Key here is that you are choosing a part of your defined population.
 Population: everyone that lives in Raleigh
 Sample: 1,000 people chosen randomly from various neighborhoods throughout Raleigh.

Population Vs. Sample

Inferential Statistics/Inference – using information from a sample to draw conclusions about

corresponding population.

Why do we use a sample and not the population?

____________________________________________________

What does inference allow us to do?

____________________________________________________

3
Sample from the Population
1) The student council at your high school decides to survey the local restaurants in order to
determine what restaurants are popular for students to attend after football games. From an
alphabetical list on White Pages.com the student council chooses 75 restaurants at random. Of
these 75 restaurants only 32 returned the questionnaire. Identify the population and the sample.
Population: all of your local restaurants
Sample: the 32 restaurants that returned the questionnaire.

2.) Verizon is trying to determine if it is worth their time and effort to extend their Fios lines
to small towns in NC. The process is expensive and time consuming and they wonder if they will
make enough money to support the installation. They go to Smallville, NC and survey the entire
town to see what proportion of people would purchase their internet services.
Population: All the small towns in NC.
Sample: The town of Smallville.

Why uses a sample?

If we want to gather information on all the high school students in NC, then why don’t we
contact ALL the high school students in NC??
1.) It’s very expensive. We might not have enough money to hire people to contact everyone!
2.) It’s time consuming. How long would it take to contact all the high school students in NC.
3.) Some people might not be able to be reached.

Why do we need good samples?

Getting a good sample is very important. Without a good sample we can’t make a reliable
conclusion about the population. Without a good sample our results are worthless!
Let’s say you go to the ice cream store, and you get a sample of mint chocolate chip ice cream before
you purchase it. One bite is a good representation of the POPULATION flavor for that stores mint
chocolate chip ice cream. However, maybe you decided you wanted moose tracks ice cream. Why is
the sample of mint chocolate chip ice cream not a good sample for the moose track?
________________________________________________________________________________
________________________________________________________________________________

4
Examples of Good Samples
Let’s take a look at the scenarios below and see what a good sample looks like.
1) You want to estimate how many students at your high school text during class.
Good Sample: You ask 5 students from every homeroom if they text during class. This gives
you someone from every grade!
Bad Sample: You ask all the students in 3 homerooms. This could end up being all the 9 th
graders and you would not end up with anyone in grades 10 – 12!

2) You want to estimate how many teenagers in the United States have had an accident while
texting and driving.
Good Sample: You randomly select 1,000 teenagers aged 16.- 18 from each state and ask if
they text and drive.
Bad Sample: You go to three high schools in your country and find all the teenagers aged
16-18 in those high schools. You ask if they text and drive.
2) This could be bad because if you are in a city, then students might not have cars. They would
ride the metro or bus. They won’t be texting and driving so it might look like less student’s
text and drive than they actually do!

Sample and Population Sorting Activity

Each week a political magazine polls a sample of 250 subscribers to their magazine. They use this
information to determine the national opinion on controversial topics. 
All readers of magazines – none
250 Subscribers to the political magazine – sample
All the subscribers of the magazine – population
In 2018 North Carolina wanted to gather information on the type of technology that students were
using at home. A survey with a $1 reward for submitting the survey was sent to students in the
Durham area. 
Students in the Durham Area – sample
All Students in North Carolina – population
All students in the United States – none
Apple purchases all of their camera parts from a certain manufacturer in the United States. There are
reports that say that the picture varies slightly from phone to phone. In order to take a look at the
quality of the supplier's product, the manufacturer ships over a sample of 15 camera parts from the
last batch to study. 
All camera parts manufactured from the entire United States – none
All camera parts from apple’s supplier – population
15 camera parts – sample

5
Samples and Inferences
In this lesson, we will continue to examine how we will use samples as representations of the population. As you
review the presentation below you will see how to gather a mean or proportion from a sample. The sample mean or
proportion that you gather will estimate the mean or proportion of the population. You will really begin to understand
why the sample must accurately reflect the population.  We want a quality sample.

Notes Two - What’s a Good Sample?

As a quick reminder it’s important to choose a good sample that represents the population. What exactly
does that mean?
That means that your sample needs to look like a mini population.
You decide that you want to study how long students at Wi-Fi High School are spending on social media.
There are 2,500 students at Wi-fi High School.
There are students that are in grades 9 - 12, some have cell phones, and some do not. Some students are
taking a large course load including AP courses and some are not. Some students participate in
extracurricular activities such as sports and band. If we want our sample to be a mini population, what
should it look like?

You decide - which one is a mini population?

Scenario 1: You go to the soccer field and after practice ask every soccer player how many hours, they
spend on social media each day.
● Yes, or no? (Circle one)
Why? ______________________________________________________
Scenario 2: You stand outside the library after every period and ask everyone that comes out how many
hours they spend on social media.
● Yes, or no? (Circle one)
Why? ______________________________________________________
Scenario 3: Since English is a required subject for all grades you go to every English class, close your eyes
and point to 5 names on the roster and ask each of those students how many hours they spend on social
media.
● Yes, or no? (Circle one)
Why? ______________________________________________________

Using a Sample to Make Inferences

Scenario 3 wins! What do we do with our information now?
You have gone to all of the English classes at Wi-fi High School, taken a sample of 5 students, and
asked how many hours they spend on social media. You then find the average amount of time from
your sample that students spend on social media. The average from your sample is 2.3 hours. What
do we do with this?

6
We are going to:
● Use this sample average or sample mean to provide an estimate to the population mean.

Population Means and Proportions

Parameter-any numerical value that comes from data collected from a population.
Population mean- the mean/average that is found from a defined population. Denoted as: Greek
symbol mu = μ

Population proportion-the proportion that is found from a defined population. Denoted as: p

Sample Means and Proportions

Statistic-any numerical value that comes from data collected from a sample.
Sample mean- the mean/average that is found from a sample. denoted as: …/x read as x bar.

Sample proportion-the proportion that is found from a sample. denoted as: … read as p hat

Think about- When we found the average amount of time that students spend on social media at
Wi-fi High School was this a statistic or parameter?
● This was a _____________________________
● This would be written as: ______________________________________
● The symbols are important because they tell you if you gathered the data from a
_____________________________
Symbols Table

Parameter Symbol Statistics Symbol

(population) (sample)

Mean μ
_____

x
Greek symbol mu

Proportion

7
Identifying Statistics and Parameters
Identify the population, the parameter, the sample, and the statistic in each of the following
settings:
A pediatrician wants to know the average weight of all 15-year-old boys, so she takes a sample of 75
patients and calculates the average weight of the 75 patients. The average weight is 176 pounds.

Population All 15-year-old boys Sample 75 15-year-old boys included in the

sample.

____
Parameter The parameter of interest is the Statistic
weight of all 15-year-old boys. X = 176
The statistic is 176 pounds, the
Mu=______ average weight of the boys in the
sample.

You Try Video! – What’s a Good Sample?

Scenario 1:
You want to study the study habits of those in your school – so you go to the library and ask 30
students how many hours of homework they do a week.
Is this a good representation of the population? This is not a good representation of the population.
Why or why not? Since you are looking at people just outside of the library, they will most likely be
studious. This will overestimate the average homework time spent on the population.
Is this using a sample mean or proportion? In this case, we would be using a sample mean as the simple
statistics.

Scenario 2
You want to know how many students at your college campus play intramural sports. You stand
outside the gym and ask 25 students if they participate in intramural sports.
Is this a good representation of the population? No, this is not a good representation of the
population.
Why or why not? Since you are asking people that already play sports, this will overestimate the
average number of students that play intramural sports.
Is this using a sample mean or proportion? In this scenario, you would be looking at a sample
proportion. The proportion of students that participate in intramural sports or

Scenario 3

8
You want to know what percentage of student’s bike to school, so you stand in the school parking
lot and ask how many people bike to school.
Is this a good representation of the population? No.
Why or why not? Since you are standing in a parking lot, this will underestimate the average number
of students that ride bikes to school because everyone is driving!
Is this using a sample mean or proportion? In this scenario, you would be looking at a sample
proportion. The proportion of students that bike to school. This would be a statistic!

Statistics or Parameter Sorting Activity

A random sample of 800 people who signed a card saying they intended to quit e-cigarettes were
contacted 11 months later. It turned out that 300 or 37.5% of the sampled individuals had not smoked an
e-cigarette over the past 11 months. What is the 37.5%? – statistic
A random sample of 800 people who signed a card saying they intended to quit e-cigarettes were
contacted 11 months later. It turned out that 300 or 37.5% of the sampled individuals had not smoked an
e-cigarette over the past 11 months. What are all the people who smoke e-cigarettes? – Parameter
A random sample of female high school students has a mean height of 62.5 inches, which is greater than
the 61-inch mean height of al female high school students. What is the 62.5 inches? -- Statistic
A random sample of female high school students has a mean height of 62.5 inches, which is greater than
the 61-inch mean height of al female high school students. What is the 61 inches? – Parameter
A large container is full of nails with mean length of 1.975 inches. This is within the specifications for
acceptance of the lengths of nails by the manufacturer. An inspector randomly chooses 150 nails from
the container that have mean lengths of 2.05 inches. Because this is outside the specified limits, the
container is mistakenly rejected. What is the 1.975? Parameter
A large container is full of nails with mean length of 1.975 inches. This is within the specifications for
acceptance of the lengths of nails by the manufacturer. An inspector randomly chooses 150 nails from
the container that have mean lengths of 2.05 inches. Because this is outside the specified limits, the
container is mistakenly rejected. What is the 2.05? statistic

Surveys vs Experiments
In this lesson you will see why you had to choose a good sample. We will explore the differences
among a sample survey, an observational study, and an experiment.  
Notes Three - Statistical Studies
In this lesson you will learn that there are three main ways to gather data:
1) Survey
2) Observational study
3) Experiment
All three of these methods are considered statistical studies.

Statistical study- a study where data is collected and the statistics (means, proportions, etc.)
are examined.

9
In this section we are going to take a look at surveys, observational studies, and experiments. Why
do we use samples to help gather information? The goal is to answer some of the questions given
below. At the end of this section, you will be able to determine if a sample, survey, or experiment is
the best way to answer the following questions.
•Do students learn better when they are listening to music or in silence?
•Do SAT scores really go up if students attend a 6-week SAT seminar?
•Do students learn better when they are sitting in rows or when they are in a horseshoe pattern?
•Do students learn better if school starts at 7:30am or 9:00 am?
•Would students in elementary schools learn better if they had a chance to have a morning snack?

Sample survey- a survey (or a questionnaire) that is given to a few members of the population.
1) Watch the wording of the question!
if you state do you like the beach – people are going to most likely say yes since you are leading
them towards “liking” the beach.
2) The person doing the interview
if a police officer asks you how often you speed, many if people will lie because they don’t want
to get into trouble with the law.
3) Where you are doing the survey
if you are doing a survey on if people prefer a book or a movie and stand outside a library! – you
are going to get more people that say they like to read – since they are at the library!

Wording of the Question

The way in which a question in a survey is worded is very important. You want the answers to be
accurate - so that means you need to word the question so that it does not guide the person being
interviewed one way or the other. Take a look below at the two examples. Which one is worded so
that the interviewer will get accurate answers?

1. On a scale of 1 - 5, with 1 being not stressful and 5 being highly stressful- how stressed do
you get by homework?
** Note here we can find the average stress level and use this information to see if the
school needs to do something to decrease stress levels!

2. Based on research done by an independent company, they estimated that 6% of trash in

today’s landfills comes from plastic water bottles. Given this information, what is your
opinion on placing a ban on plastic water bottles in the United States? Yes.
We can use surveys to answer questions that we might have, the next thing we want to take a
look at is observational studies.
Observational studies- a method of collecting information about a population by observing
subject without imposing treatment (i.e., examining the health effects of smoking).

10
This would be when you would:
1.) Ask questions to gather information – example ask you friends how many hours of sleep
they get. You then find the average number of hours your friends sleep to determine if high
school students are getting enough sleep.
2.) Observe a behavior and record what is happening. Let’s say that you study traffic patterns
–you sit at a traffic intersection and see how many cars run a red light for a week. Then
you find the average number of cars that run that red light in a day.
3.) We would use observational studies when we can’t give a treatment to someone. For
example, if we want to examine the effects of smoking, we can’t force people to smoke. If we
want to check out how pesticides on food affect your health later in life, we can’t force
people to each pesticide.

After doing a very large study of coffee drinkers, the Arizona Daily Star wrote an article titled
“Coffee Buzz: Study Finds Java Drinkers Live Longer”.

● In this study they examined 400,000 people and the amount of coffee they drank per day.

● They took into consideration the amount people smoked, drank alcohol, and exercised.

● The study found that that coffee drinkers live a little longer than non-drinkers,
regardless of whether they drink regular or decaf.

Explain how this is an observational study.

This is an observational study because they asked people how many cups of coffee, they drank each
day. They did not force people to drink or not drink coffee.

Identifying Vocabulary –
Take a look at the below experiment and identify the vocabulary listed in the table.
Mrs. Price wants to determine whether or not caffeine affects the pulse rate of students in
her classroom. She randomly assigns 15 students to drink caffeinated coke and the other
half to drink decaffeinated coke. She puts the two different kinds of coke in the same type
of glasses, so they look exactly the same. The students will not know whether they are
drinking caffeinated or non-caffeinated coke. She measures their pulse rate before they
drink the coke and then 15 minutes after.

Placebo This is the decaffeinated coke. It looks exactly like the caffeine but
does not have any caffeine in it.

Treatment The caffeinated coke.

11
 Control In this case, the placebo is the control. The decaffeinated coke – is
what they will compare the treatment to.

What are they comparing? In the end, they are comparing the pulse rates of the students that
drank the coke with caffeine and the coke without caffeine.

Experimental Studies
The next type of statistical study we can do is an experiment. Before we dig into experiments, let’s take
a look at some of the vocabulary for experiments.
Experiments- a method of applying a treatment to a group and recording the effects.
Placebo – it is a “fake” treatment. It looks like the treatment, smells like the treatment, so people in
the experiment think they are getting the treatment! Control.
Treatment Group – the group in an experiment that receives the treatment.
Control Group – something to compare the treatment to. Everyone receiving the control will be
placed in the control group.
- a control group is given a placebo/or control and then the reactions between the treatment
group(s) and the control group are compared.
.
Why would we use an experiment instead of an observational study?

● We want to use an experiment when we want to determine cause and effect.

What does this mean? It means that we want to see if a new drug causes less headaches.
3) Does caffeine cause your pulse to go up? –

● You can only establish cause and effect through experiments NOT observational studies. –
treatment control

Take a look at the experiment below and identify the vocabulary listed in the table.
Dr. Healthnut and her colleagues gathered 150 obese adults at the Heart Hospital in Raleigh, NC.
She wanted to determine whether or not eating a low carbohydrate diet would help to decrease
cholesterol levels and help to lose weight. Of the 150 adults she assigned one group to each a low
carb diet and the other a standard diet. At the one year and 2-year mark she measured the
participants change in cholesterol levels and weight.

Placebo -- NONE
fake looks exactly treatment

12
 Treatment Low-Carb diet

Control – comparing treatment Standard diet

What are they comparing? --- The weight and cholesterol levels of the participants in the
end of exp. control group to the weight and cholesterol levels of the
participants in the treatment group.

When would you use a sample survey, observational study, or experiment?

● You wanted to gather more information if a toll should be added to highway 66.
Sample survey – you want to gather information to help make a prediction.
● How does vaping affect lungs in 20 years?
Observational study – you don’t want to force people to smoke, but you can observe the
effects.
● Will providing students with 3 or 6 more minutes between classes allow them to get to
class on time?
Experiment – you can provide a treatment of 3 minutes and 6 minutes and observe the
results.

Which is best
Would the following questions best be answered as an observational study, sample survey, or
experiment?
1. Do students learn better when they are listening to music or in silence? –
 Experiment
 Give half of the student’s music to listen to in class and make the other half work in
silence. Give a test on the material at the end of the class and compare to see which group
did better.
2. Do SAT scores really go up if students attend a 6-week SAT seminar? – experiment
 Observational Study or Experiment. 
 For an observational study you could compare the results of the SATs from a group of
students that took a 6-week SAT seminar and a group that did not. 
 For an experiment you can have half the students go to the 6-week SAT seminar and a group
of students do not attend. From there you could compare the SAT scores of the two groups.
3. Do students learn better when they are sitting in rows or when they are in a horseshoe
pattern? – experiment
 Experiment. 
 Place one class in rows and one class in a horseshoe pattern. At the end of class compare
the students results on an assignment to see which class understood the material
better.

13
4. Do students learn better if school starts at 7:30am or 9:00am? – observational study
 Experiment. 
 Start school for half the year at 7:30 am and complete the other half of the school year
at 9:00 am. Compare the grades at the end of the year.
5. Would students in elementary schools learn better if they had a chance to have a morning
snack?
 Observational study.
 Find an elementary school that does not provide a morning snack and another elementary
school that does provide a morning snack and compare student performance.

You Try!
Take a look at the three scenarios below. Explain why each of the following are misleading
questions.

1) Indicate whether you agree or disagree with the following statement:

“I feel welcomed by the coaches and athletes on my first day at lacrosse camp.”
The statement “I feel welcome” leads to state that you feel welcomed even if you did not.
The question is leading you in a certain direction .

2) During the last school year, the administration for a local school district wanted to see if
the community supported an increase in taxes to increase pay raises for teacher and more
technology for students. The survey was sent to every household in the school district.
One of the questions they asked was: “Even if you are not currently involved in the school,
do you think that we should raise taxes in order to increase funding for our schools?” 66%
of the respondents answered no.
The statement in bold indicates that since you have no stake in the school, so why would
you want to spend your money on the school.

3) A local expert psychologist was looking at the effects of TV on younger children. She
personally went t 35 homes in the Durham area with preschool aged children. She asked
the following question: “Do you agree or disagree with the fact that if your child watches
more than 3 hours of TV per day then they will be less likely to sit still in kindergarten?”
The question is actually worded pretty well! However, since it is coming from an expert in
psychology, people Night tend to answer yes. -TV did not affect how likely they will slot
in kindergarten.

14
 Question 1
Your family has decided to open up an Air B and B in their spare bedroom and bathroom. You want
to find out how the people staying at your Air B and B like their stay. Write a survey question that
will appropriately answer the question.
 Make sure to avoid questions that indicate a direction.
 Make sure if you have a scale that it goes in order fro not satisfied to satisfied.

Survey question:

Question 2
Your community has decided to impose a curfew on drivers under 18. You believe that most parents
do not support this curfew. Write a survey question to find out what percentage of parents with
teenagers would support this curfew. Be careful not to include your opinion in the question!
 Make sure when you ask the question that you don’t state – most parents don’t agree
with the curfew.
 An example would be “Currently, there is no curfew for teenagers under the age of 18.
Would you or would you not support a law that imposes a 11:00PM curfew for teenage
drivers under the age of 18?

Survey question:

Finding the Sample Mean and Sample Proportion

Question 1
Jack Frost High School is trying to decide on whether or not they should remove the soda machine
from the cafeteria. The principal has stated that 75% of the students would support this. You
think this is crazy. You take a random sample of 250 students at your school and 75 say they
support taking out the soda machine. What is the sample proportion?
Outcome from the survey: _________ The 75____________________
Sample Proportion: _______________________

15
Question 2
You are getting ready to figure out your monthly electric bill. The electric company recommends
that you budget $75.00 a month based on the building’s electric use. You think this is a lot! You
randomly sample 5 people in your apartment building with the same floor plan as yours and ask
what their monthly electric bill totals. They say $67, $35, $100, $72, $57. What is the sample mean?
Readings were: $67, $35, $100, $72, $57.
Sample Mean: _________________________________________________

OBSERVATIONAL STUDY OR EXPERIMENT SCENARIOS

Read each scenario and decide whether pr not each scenario is an observational study or an experiment.

Aaron is a video game designer. He and all his friends prefer Monster Energy Drink over drinking coffee in the
morning. (Caffeine in Monster has twice as much caffeine as one cup of coffee). Aaron decides that most
game designers prefer an energy drink over coffee.
Is the example above an experiment or observational study? Do you think it’s valid? Why or why not?
4) It is an observational study because Aaron is basing his conclusion on what he observes people in his
office drinking. He has not imposed a treatment on anyone or forced them to drink Monster. It’s not a
good conclusion because the sample only includes those in his office - not video game designers
anywhere else.

Investigators wanted to C of there was a connection between the use of cell phones and the risk of brain
cancer. One study observed 672 people with brain cancer. The investigators separated the people by age, sex,
and race. The researcher then randomly chose 672 people without brain cancer and separated them by age,
race, and sex. In all six groups they calculated the average number of hours that each group spent on the cell
phone and compared the corresponding groups. Their result was: “Our data suggests that the use of hand-
held cellular phones is not associated with the risk of brain cancer.”
Is the example above an experiment or observational study? Do you think it’s valid? Why or why not?
5) This is an observational study because they did not force the people in the study to use a cell phone.
They took people that already had brain cancer and people that did not and observed their cell phone
use. They did not impose any treatment.

The NCAA is interested in checking to see if a new protein powder will help to build muscle strength in
football players. They take a random sample of 700 college football players and have them do the bench
press. They randomly assign half of the football players to drink a protein shake for 30 days hey test their
bench press weight again to see if their strength was improved.
Is the example above an experiment or observational study? Do you think it’s valid? Why or why not?
6) This was an experiment since they imposed a treatment. The treatment was the protein powder that
they gave half of the football players, and the control group was the group that did not get any
protein powder.

16
 You are checking to see if you can get grass to grow in your shady backyard. You have decided to look at two
types of fertilizer to see if any of them help the grass to grow. You plant grass in three different places. In one
spot you let the grass grow in the soil that’s there, in the second place you apply Scott’s fertilizer, and in the
third spot you use University of Virginia’s special shady grass fertilizer. After giving all grass equal amounts of
water, at the end of 30 days you see which area grew the most grass.
Is the example above an experiment or observational study? Do you think it’s valid? Why or why not?
7) This is an experiment because there was a treatment imposed on the grass. Control group was the
group with no fertilizer, the other two treatments were Scott’s fertilizer and UVA’s special shady grass
fertilizer. 

INFERENCE FROM STATISTICAL STUDIES

In the last three sections, we have looked at what makes a good sample and how we use the sample to make
predictions about the populations. In this section, you will take a look at various scenarios and decide whether
or not you can make an inference about the population based on the data gathered

Notes Four - Scenarios

During this module you have learned about samples and how we use samples to make inferences
about population.
In the scenarios you will be

1) deciding whether or not it’s an experiment or observation,

2) determining if we can make a conclusion and

3) make a decision on whether or not we can make inferences about the population.

CONCLUSION? INFERENCE?
How will we know if we can make a conclusion?
 make sure the sample was randomly chosen. If it was then you can make inference to the
larger population. If it was not, then the conclusion can ONLY be made to those in the
study.
 In order to establish cause and effect (the treatment made a difference) the treatment must
be randomly assigned to the treatment and control groups.
Random sample  conclusion to population
Random assignment 
Scenario 1
Does having a Bluetooth in the car reading text messages decrease driver distractions from the
road?

● Investigators randomly recruited 80 subjects to participate in an experiment.

17
 ● They had a driving simulator that was equipped with Bluetooth that read text messages
every time they were received.
● The researchers measured how many times the drivers took their eyes off the road.

● 40 participants completed the drive with a cell phone in the car while the other 40
participants completed the drive had their incoming text messages read to them via
Bluetooth.
● The route, driving conditions, and traffic flow was the same for both groups.
At the end of the drive the participants that had their text messages read to them looked away
from the road an average of 10 times. The participants that had their cell phone with them but did
not have their text messages read to them looked away from the road an average of 25 times.
Let’s answer some questions regarding the study!

1) What conclusions can you make from the study?

Overall, even though we can make inference to a larger population – since we can’t establish
that having text messages read aloud caused the drivers to look at their cell phone less there
are no valid conclusions we can make from the study.
8) Investigators randomly recruited 80 subjects to participate in an experiment
o Since they randomly recruited 80 subjects – they can make a conclusion about their
study to the larger population. The subjects were randomly chosen.
9) 40 participants completed the drive with a cell phone in the car while the other 40
participants completed the drive while having their incoming text messages read to them
via Bluetooth
o Notice here that they were not randomly assigned to the cars that read the text
messages to them or not. Therefore, we cannot establish cause and effect. We
cannot say that having the text messages read to them caused them to look at their
cell phone less.

2) Was this an observational study, sample survey, or experiment?

Experiment – they applied a treatment of reading text messages in the car.
 40 participants completed the drive with a cell phone in the car while the other 40
participants completed the drive while having their incoming text messages read to them
via Bluetooth.
Think about your answers before we move onto the next slides!

Scenario 2
Does the size of the class matter in grade performance for high school math in NC?
Will students do better on their End of Course Exams if they are in smaller classes?

● Researchers randomly chose and observed two Math III courses at Langley High School,

18
 one with 15 students and the other with 30 students.

● At the end of the school year the average on the EOC for the class of size 15 was an 85% and
the class with size 30 was 75%.
Was this an observational study, sample survey, or experiment?
Observational study – there was not treatment imposed.
What conclusion can be made from the study?
Since the sample was randomly chosen, inference can be made to the larger population.
This is not an experiment – they researched classes already in session, so we do not need
to worry about random assignment.
We can conclude that the smaller class size does help improve EOC scores. However, we
may want to think about choosing more/different high schools to study.

You Try!
With the growing use of technology by students in the classroom, many educators believe that
students rely solely on technology to solve even the simplest of problems. Their ability to analyze
and solve in depth problems has decreased significantly. Anything longer than the length of a
tweet means that students often lose interest in the task at hand.
Researchers decided to test whether or not banning cell phones increased math scores and
critical thinking skills. In the state of North Carolina, they randomly chose 15 high schools to
test. At those 15 high schools the researchers allowed half of the math classes to have cell phones
and computers in class and in the other half of the math classes the cell phones were collected at
the beginning of class. The researchers then gave a pretest to all math classes at the 15 high
schools to test the students critical thinking skills. The teachers at the 15 high schools taught a
set curriculum for 3 weeks. At the end of three weeks the students were given the same test again.
Researchers compared the results from the classes that were allowed technology versus those
that were not allowed technology. They found that at the end of three weeks the proportion of
students that were able to answer the critical thinking questions in the non-technology
classrooms was 66% and the proportion of students in the technology classrooms was 60%. Did
taking away the students’ cell phones help to increase the students critical thinking skills?

What conclusion can be made given the study? --- What is your conclusion?
Let’s think about this:

1) The sample was 15 high schools in NC – therefore the conclusion can only be made for math
students in NC high schools.

2) The proportion of students that got the critical thinking questions correct from banned
technology and technology was 66% and 60%. This does not seem to be a huge difference.

19
 3) Since there was not a large difference in the proportions, based on this information, we cannot
make a conclusion that taking away technology made a difference in critical thinking skills.

Extra Notes:
________________________________________________________________
________________________________________________________________

20