Unit 8 – Hypothesis Testing – Part 1 Name:_____________________Block:____

Lesson 1 – Hypothesis Testing: The Basics

Opening Activity
• Let’s start off with a small activity: I need a volunteer!
• We are going to see how well ____________ can play trashketball! They will make 10 throws of this
balled up trash from a spot on the floor into the garbage can.
• ____________, how many do you believe you can make out of 10? _______________
• Class, do you believe he will make more or less than this? ___________________
We are going to test this claim!

Claim one: ____________ Claim two: ____________

These claims are called hypotheses and the test to determine which one we will go with is a significance test.

Claim one: ____________ This claim is called our null hypothesis: _______
This can be a claim made by someone/something or a statement of “no difference”

Claim two: ____________ This claim is called our alternative hypothesis, _______
This is the claim we are trying to find evidence for.

Let’s have them shoot. What was their sample proportion?

Was it exactly the claimed proportion? Do we expect it to be?

Can they have gotten this sample proportion and their claim still be correct?

How can we decide if the difference is chance variation or if their original claim is incorrect?

Page 1
One-Sided Hypothesis Test

• In significance testing, we set up two hypotheses: Example 1

o The __________ hypothesis is an original claim, and is always
the population proportion equal to a value
o The ____________________ hypothesis is the claim we are
trying to find evidence for Example 2
• In a one-sided hypothesis test, the inequalities are either ___________
the value in our null hypothesis
• ________ is called our claimed proportion

Notice some similarities:

● The null hypothesis is always ________________ a value
● The alternative is above or below that __________________
● Our hypotheses are based on our population parameter ______

Two-Sided Hypotheses Test

• In a two-sided test, we are looking for evidence against the claim, but we Example
are unsure about __________________________________
• Hypotheses are developed before a study is conducted, so after we
gather 𝑝̂ and we know which direction the difference is, we
_______________________ and change our alternative hypothesis

A hypothesis test asks the question:

How likely is it to get a result like this just by chance, when the null hypothesis is true?

In our opening activity, if a student said they were a 80% trashketball shooter, it is possible for them to shoot
50% in a single game?

This is a conditional probability question: ____________________________________________

If we find the conditional probability and it is small, this is good evidence ____________________________
The probability of the event occurring given an original claim is called the _______________

Note: A Hypothesis Test is also called a Significance Test and the two terms can be used interchangeably.
Page 2
P-Value
• The ______________________, computed assuming H0 is true, that the statistic 𝑝̂ would take a value as
extreme or more extreme than the one actually observed, in the direction specified by Ha
• We compute this probability by using our sample statistic 𝑝̂ and the value in our null hypothesis, p0
• We will be calculating z-scores to determine the difference between what we observed (____________)
and what our claim is (___________________)
• We will then calculate the probability we would get this value or more extreme (in the direction of the
alternative) and make a decision:

if the probability is large enough, we could if the probability is low enough, we could

Significance Level
• How low is a low enough probability? This is determined by our _________________________
• This significance level is represented by the symbol ___________________

If our p-value is smaller than α, ____________________ we say that the

• result of the study is _____________________________________
• The probability of this event occurring by chance, if H0 is true, is so low, we say that the difference is
NOT due to sampling variability
• In this case, ___________________________ and conclude that _____________________________
_____________________________________

If our p-value is larger than α, ____________________ we say that the

• result of the study is _____________________________________
• The difference between 𝑝̂ and p0 can be attributed to sampling variability
• In this case, ___________________________ and conclude that _____________________________
_____________________________________
• This does NOT mean we “accept H0”!! (NEVER SAY THIS). All we showed was that we do not have
enough evidence against the claim. We did not prove the claim.

The usual significance levels are

• 1% ( = ______)
• 5% ( = ______)
• 10% ( = ______)
Page 3
Unit 8 – Hypothesis Testing Name: ________________________
HW 1 – Hypothesis Testing: The Basics

1) For the following, identify the null and alternative hypotheses.

a) A company claims that 70% of its customers are b) A university states that 60% of its students pass
satisfied with their service. A disgruntled customer their exams on the first attempt. A teacher believes
believes this is an overstatement, and wants to the actual proportion of students who pass on the first
conduct a survey to test this claim. A random sample attempt is much higher. They take a sample of 150
of 200 customers is taken, and it is found that 130 of students who take a first quarter exam, and 80 pass
them are satisfied. on their first try.

c) You state health department claims that 40% of the d) A manufacturer claims that only 10% of its
population exercise regularly. A local health official products are defective. A quality control specialist
wants to test this claim in their town. They send out a wants to test this to make sure that proportion is not
survey to 500 residents, and it is found that 220 of higher than the claim. In a sample of 300 products,
them exercise regularly. they found 35 are found to be defective.

e) A politician claims that 55% of voters in their f) A technology company claims that 25% of its users
district support their policies. A recent smear upgrade to the premium version of their software
campaign from the other political party has them after the free trial. A watchdog organization wants to
worried that this support might be lowered. Their investigate this claim to see if they find evidence that
team takes a poll of 1000 voters, and find that 525 of it is not true. In a sample of 400 users who tried the
them support the policies. free trial, 110 upgraded to the premium version.

2) For the following, use a significance level of  = 0.05 and the given information to state a conclusion about
the null hypothesis (reject or fail to reject) and if this is convincing evidence of the alternative hypothesis.

a) Claim: More than 60% of adults would erase all of their personal information online if they could. The
hypothesis test results in a P-value of 0.122.

Page 4
b) Claim: The mean pulse (in beats per minute) of health adult females is 75 bpm. The hypothesis test results in
a P-value of 0.034.

c) Claim: Less than 80% of seniors at the high school pass their AP exams. The hypothesis test results in a P-
value of 0.083.

d) Claim: The proportion of left-handed people in a city is different than the national average of 10%. The
hypothesis test results in a P-value of 0.0012.

3) For the following questions, write T for true and F for false for the given statement. If it is false, in the space
to the right, write down how it should be changed to make it true.
When conducting a significance test for proportions, the probability
a)
computed assuming H0 is true.

If the p-value is larger than α, we reject H0 and conclude there is

b)
convincing evidence in favor of the alternative hypothesis.

When the p-value is smaller than α, we always say that the result of
c)
the study is statistically significant.

We should always check the 10% condition to see if our

d)
independence condition is satisfied.

When the p-value is larger than α, the difference between 𝑝̂ and p0 is

e)
always attributed to sampling variability.

We check the large counts condition by using 𝑛𝑝̂ ≥ 10 and

f)
𝑛(1 − 𝑝̂ ) ≥ 10.

Saying “fail to reject the null hypothesis” and “accept the null
g)
hypothesis” are the same thing.

The probability of an event occurring by chance, if Ha is true, is

h)
always low when the result of the study is statistically significant.

Page 5
Lesson 2 – Hypothesis Testing for a Proportion

Conducting a Hypothesis Test for a Proportion

Step 1: Set up our hypotheses

• Our null hypothesis is always equal to a __________________________
• Our alternative hypothesis is what we suspect is incorrect about the claim

One-Sided Hypothesis Test Two-Sided Hypothesis Test

𝐻0 : 𝑝 = 𝑝0 𝐻0 : 𝑝 = 𝑝0 𝐻0 : 𝑝 = 𝑝0
𝐻𝐴 : 𝑝 < 𝑝0 𝐻𝐴 : 𝑝 > 𝑝0 𝐻𝐴 : 𝑝 ≠ 𝑝0

Step 2: Are the conditions met to perform inference (can we run the test)?
• Random (Random Sample or Random Assignment)
• Independent (When sampling, check 10% condition)
• Normal (Check large counts condition with 𝑛𝑝0 ≥ 10 and 𝑛(1 − 𝑝0 ) ≥ 10)

Step 3: Calculate your test statistic and p-value.

• Calculate your test statistic:

• Find your p-value by calculating the probability of getting a z-statistic this extreme or more extreme in
the direction of the alternative hypothesis. (On TI-84, use normalcdf)

𝐻𝐴 : 𝑝 < 𝑝0 normalcdf(L: ______, U: ______, : ___, : ___)

𝐻𝐴 : 𝑝 > 𝑝0 normalcdf(L: ______, U: ______, : ___, : ___)

2 * normalcdf(L: _____, U: _____, : 0, : 1)

𝐻𝐴 : 𝑝 ≠ 𝑝0 The boundaries depend on the direction of your
The inequality will be towards the tail inequality

Page 6
Step 4: Conclude the problem by reporting your statistical findings and what this means in the context of the
problem.
• Statistical Findings: Because our p-value of <value> is <bigger/smaller> than our significance level of
<level>, we <fail to reject/reject> the null hypothesis.
• Context of the Problem: We <don’t/do> have convincing evidence that the population proportion
<context of what this means in the problem>

Important Significance Test Notes

• The p-value is a _______________________; it assumes the null hypothesis is true. Assuming that H0 is
true, the p-value measures how likely it is to get evidence for Ha (as “large” or “larger” in the direction
of the alternative) by chance alone.
• For a hypothesis test, we check if ____________________________________ and for confidence
intervals we check __________________________________________
• When there is not convincing evidence to reject H0, is does not mean there is evidence that H0 is true or
Ha is false.

Example: For the following problems, find the p-value and write a statistical conclusion using a
significance level of  = 0.05.

a) z = −0.75 and was found with the alternative hypothesis of p < 0.25

b) z = 2.14 and was found with the alternative hypothesis of p > 0.55

c) z = 1.78 and was found with an alternative hypothesis of p  0.37

Page 7
Example: One Sample Z Test for a Proportion
Allan reads a report that says 80% of US high school students have a computer at home. He believes the
proportion is smaller than 0.8 at our school. Allan chooses an SRS of 60 students and records whether they have
a computer at home. He wants to test this at a 5% significance level. In his sample, 41 of the students have a
computer at home. Is this evidence that the proportion of students who have a computer at home at our school is
less than the national proportion?

Step 1: Set-up our hypotheses.

Step 2: Are the conditions met to perform inference?

Step 3: Calculate our test statistic and p-value

Step 4: Conclude the problem by reporting your statistical findings and what this means in the context of the
problem.

Interpret the p-value in the context of the problem.

Page 8
Activity: Hypothesis Testing with ESP

Introduction

Extrasensory perception, often abbreviated as ESP, is the ability to gain information through means other than
the known human senses. It is sometimes referred to as the "sixth sense" and includes abilities such as telepathy
(mind-to-mind communication), clairvoyance (perceiving distant events), and precognition (predicting future
events).

While some people believe in ESP, scientific research has not conclusively proven its existence. Experiments
and studies are often conducted to test claims of ESP, but results are usually inconclusive or attributed to
chance. ESP experiments involve hypothesis testing to determine if the observed abilities of individuals are
statistically significant or if they could occur by random chance.

In today's activity, you will perform a simple ESP experiment to test whether someone in your group can
identify shapes (circle or square) without seeing them. By analyzing the results using hypothesis testing, you
will determine if the subject's ability to identify the shapes is statistically significant or likely due to random
guessing.

Experiment Setup

• Select one member of the group as the subject.

• Draw a circle on one piece of paper and a square on another piece of the same size.
• Repeat the experiment 20 times:
o Randomly select one of the shapes and place it in the subject's hand behind their back so it
cannot be seen.
o Ask the subject to identify the shape without seeing it.
o Record whether the response was correct or incorrect in the table below.
• After the experiment has been repeated 20 times, show the subject their result and let them copy down
the data. Then answer the questions that follow.

Data Collection

Trial Shape Chosen (circle/square) Subject’s Response (circle/square) Correct (yes/no)

1
2
3
4
5
6
7
8
9

Page 9
 10
11
12
13
14
15
16
17
18
19
20

Proportion of Responses Correct: ______/20

Inference Questions

1) State the hypotheses for this experiment.

2) Are the conditions met to perform inference? Explain.

3) Calculate the z test statistic. Show your work!

Page 10
4) What is the p-value associated with your test statistic?

5) Using a 10% level of significance, interpret this result in the context of the problem.

6) Using a 5% level of significance, interpret this result in the context of the problem.

Page 11
HW 2 – Hypothesis Testing for Proportions

1) Give the following information, find the z-score, p-value, and write a conclusion without context.

Conclusion without
Given z-score p-value
Context

𝑝̂ = 0.4 𝑝0 = 0.25
n = 50  = 0.01
Alternative: >

𝑝̂ = 0.5 𝑝0 = 0.61
n = 45  = 0.05
Alternative: <

𝑝̂ = 0.3 𝑝0 = 0.45
n = 55  = 0.05
Alternative: 

𝑝̂ = 0.84 𝑝0 = 0.80
n = 450  = 0.05
Alternative: 

2) Determine if the following statements are true (T) or false (F)

a) It is helpful to examine your data before deciding whether to use a one-sided or two-sided hypothesis
test.
b) If the p-value is 0.05, the probability that the null hypothesis is correct is 0.05.
c) A two-sided hypothesis test is used when we are unsure about the direction of the difference from the
null hypothesis
d) The larger the p-value, the more evidence there is against the null hypothesis.
e) If the p-value is small enough, we can conclude that the alternative hypothesis is true.

Page 12
3) A company claims that 70% of its customers are satisfied with their service. A disgruntled customer believes
this is an overstatement, and wants to conduct a survey to test this claim. A random sample of 200 customers is
taken, and it is found that 130 of them are satisfied. Are the results significance at the 5% significance level?

Step 1: Hypotheses Step 2: Conditions

Random:

Independent:

Normal:

Step 3: Calculate Z Statistic and P-value Step 4: Conclusion in Context

Sample proportion:

z-score and p-value calculation:

Page 13
4) A technology company claims that 25% of its users upgrade to the premium version of their software after
the free trial. A watchdog organization wants to investigate this claim to see if they find evidence that it is not
true. In a random sample of 400 users who tried the free trial, 120 upgraded to the premium version. Is this
evidence that the company’s claim is incorrect? Use a 5% level of significance.

Step 1: Hypotheses Step 2: Conditions

Random:

Independent:

Normal:

Step 3: Calculate Z Statistic and P-value Step 4: Conclusion in Context

Sample proportion:

z-score and p-value calculation:

Page 14
Lesson 3 – Hypothesis Testing for Means

Conducting a Hypothesis Test for a Mean

Step 1: Set up our hypotheses

• Our null hypothesis is always equal to a ____________________
• Our alternative hypothesis is what we suspect is incorrect about the claim

One-Sided Hypothesis Test Two-Sided Hypothesis Test

𝐻0 : 𝜇 = 𝜇0 𝐻0 : 𝜇 = 𝜇0 𝐻0 : 𝜇 = 𝜇0
𝐻𝐴 : 𝜇 < 𝜇0 𝐻𝐴 : 𝜇 > 𝜇0 𝐻𝐴 : 𝜇 ≠ 𝜇0

Step 2: Are the conditions met to perform inference (can we run the test)?
• Random (Random Sample or Random Assignment)
• Independent (When sampling, check 10% condition)
• Normal:
o ________, the _________ tells us the sampling distribution will be approximately normal
o ________, we have to either:
▪ Have the original population be approx normally distributed
▪ Graph the sample data to make sure there are no outliers or strong skewness

Step 3: Calculate your test statistic and p-value.

• Calculate your test statistic:

• where t has n – 1 degrees of freedom

𝑠
• 𝑥̅ and s come from your sample data, and because of that, the formula represents your standard error,
√𝑛
and this is why a t-test statistic is needed
• if you know  (which remember, you rarely do), we would perform a __________________ for a mean:

𝑥̅ − 𝜇0
𝑧= 
√𝑛

• Once we have our t-test statistic, we use tcdf(Lower Boundary, Upper Boundary, df) to find our p-value.

Step 4: Conclude the problem by reporting your statistical findings and what this means in the context of the
problem.
• When _______________, we can reject the null hypothesis in favor of the alternative hypothesis
• When _______________, we fail to reject the null hypothesis and say we do not have evidence of the
alternative hypothesis
Page 15
 • Statistical Findings: Because our p-value of <value> is <bigger/smaller> than our significance level of
<level>, we <fail to reject/reject> the null hypothesis.
• Context of the Problem: We <don’t/do> have convincing evidence that the population proportion
<context of what this means in the problem>

Example: One Sample T Test for a Mean

Your AP Statistics teacher believes that the average number of minutes their students spend on homework every
night is greater than 35 minutes. She randomly selects a sample of 10 students and records the number of
minutes they spent on homework last night. The data are below.

32 28 48 33 34 59 62 45 38 41

Is there convincing evidence, at the 10% significance level, of your teacher’s beliefs?

Step 1: Set up our hypotheses

Step 2: Are the conditions met to perform inference?

Step 3: Calculate your test statistic and p-value.

Step 4: Conclude the problem by reporting your statistical findings and what this means in the context of the
problem.

Page 16
HW 3 – Hypothesis Testing for Means

1) Give the following information, find the z-score, p-value, and write a conclusion without context.

Conclusion without
Given t-statistic p-value
Context

𝑥̅ = 10.5 s = 2.85
𝜇0 = 8.5
n = 10  = 0.05
Alternative: >

𝑥̅ = 157 s = 25.78
𝜇0 = 170
n = 15  = 0.01
Alternative: <

𝑥̅ = 32.6 s = 9.8
𝜇0 = 35
n = 24  = 0.10
Alternative: 

𝑥̅ = 26.3 s = 6.5
𝜇0 = 24
n = 32  = 0.10
Alternative: 

________ 2) What is the average teacher salary in the United States? One district claims that it is $43,500 and
will therefore set the salary level based on that. The teacher’s union says that it is incorrect and the actual
income is higher and the district should set their salary level higher. The teacher’s union selects an SRS of 15
teachers from around the US and finds that the mean salary is $45,000 with a sample standard deviation of
$2,800. Assuming that the distribution of all teachers’ salaries is approximately normally distributed, what is the
value of the t-test statistic and the P-value for a test of the hypothesis 𝐻0 : 𝜇 = 43,500 against 𝐻𝑎 : 𝜇 > 43,500?

(A) t = 2.075 0.01 < p < 0.05

(B) t = 2.075 0.001 < p < 0.01
(C) t = –2.075 0.01 < p < 0.05
(D) t = –2.075 0.001 < p < 0.01
Page 17
3) Functional Fitness claims that its new exercise program helps their members lose 5 pounds in the first month.
To test this claim, you select a random sample of 13 members who completed the program and record their
weight loss (in pounds) as follows:

5.5 5.2 6 5.5 6.8 4 4.2 5.5 6.2 5.1 4.8 6.6 3.9

Is there evidence, at the 10% significance level, that this new exercise program helps members lose more than 5
pounds in their first month?

Step 1: Hypotheses

Step 2: Conditions

Random:

Independent:

Normal:

Step 3: Calculate T Statistic and P-value

Sample Statistics:

𝑥̅ = _______ 𝑠 = _______ n = ______ df = ______

z-score and p-value calculation:

Step 4: Conclusion in Context

Page 18
Unit 9 – Hypothesis Testing Name: ________________________
In Class Practice for L1 – L3

1) For the following situations, identify the null and alternative hypotheses.
a) A city council claims that over 60% of its residents b) A nutritionist believes that a new diet plan will
support constructing a new Amtrack railway line lower the average weight of participants. The average
through town. To verify this claim, a local newspaper weight of participants before starting the diet is 200
conducts a poll of 500 randomly selected residents pounds. After following the diet for three months, the
and finds that 320 support the initiative. average weight of a sample of 40 participants is 195
pounds with a standard deviation of 14.

c) A coffee shop owner wants to know if the average d) A university states that at least 75% of its
amount of coffee sold per day has changed from last graduates find employment within six months of
year. Last year, the average daily coffee sales were graduation. To test this claim, a survey is conducted
150 cups. The owner randomly selects 30 days from among 100 recent graduates, and it is found that 78
the current year and finds that the average daily are employed within six months.
coffee sales are 160 cups with a standard deviation of
25 cups.

2) For the situations above, assume that all conditions necessary for inference have been met. Find the
appropriate test statistics, p-value, and write a conclusion in the context of the problem, using a significance
level of 10%. Show your work!

a) A city council claims that over 60% of its residents b) A nutritionist believes that a new diet plan will
support constructing a new Amtrack railway line lower the average weight of participants. The average
through town. To verify this claim, a local newspaper weight of participants before starting the diet is 200
conducts a poll of 500 randomly selected residents pounds. After following the diet for three months, the
and finds that 320 support the initiative. average weight of a sample of 40 participants is 195
pounds with a standard deviation of 14.

Page 19
 c) A coffee shop owner wants to know if the average d) A university states that at least 75% of its
amount of coffee sold per day has changed from last graduates find employment within six months of
year. Last year, the average daily coffee sales were graduation. To test this claim, a survey is conducted
150 cups. The owner randomly selects 30 days from among 100 recent graduates, and it is found that 78
the current year and finds that the average daily are employed within six months.
coffee sales are 160 cups with a standard deviation of
25 cups.

3) True or False? Write T for true and F for false on the line corresponding to the statement.

a) The conclusion of “fail to reject the null hypothesis” has exactly the same meaning as “accept the
null hypothesis”.
b) When conducting a hypothesis test with a convenience sample, we can still satisfy the random
condition by using a large enough sample size.
c) If we repeat the same hypothesis test with different random samples of the same size, the statistical
conclusions will all be the same.
d) A significance test is designed to measure the strength of evidence again the null hypothesis.
e) A one proportion z-test requires the sample size to be large (n  30) to ensure the sampling
distribution we use is approximately normal.
f) A large p-value is evidence for the alternative hypothesis.
g) If a sample is large enough, you do not need to have a random sample
h) If you graph sample data for a one sample T test for means and there are no outliers or strong
skewness, that means you can assume the data comes from an approximately normal population distribution.

Page 20
______ 4) A school district has past data indicating that given an incoming freshman class, 94% of those
students will end up graduating on time. A teacher in your district is worried that, because of the pandemic, that
value will be less than 94% for the upcoming graduating class. With H0: p = 0.94 and Ha: p < 0.94, the p-value
of the test is 0.028. Which of the following is a correct interpretation of this p-value?

(A) If less than 94% of students will graduate, the probability of obtaining a sample proportion equal to or
larger than the one the teacher got is 0.028.
(B) If less than 94% of students will graduate, the probability of obtaining a sample proportion as small or
smaller than what the teacher got is 0.028.
(C) If 94% of students will graduate, the probability of obtaining a sample proportion as small or smaller than
what the teacher got is 0.028.
(D) If 94% of students will graduate, the probability of obtaining a population proportion less than 0.94 is
0.028.

______ 5) Recidivism happens when individuals who have been previously convicted of a crime, commit
additional offenses after their release from prison. The recidivism rate is often used as a measure to assess the
effectiveness of rehabilitation programs. In 2023, the national recidivism rate is estimated to be 37.1%. A social
justice activist believes that the true recidivism rate is lower at the prison near him. He selects a random sample
of 300 released inmates and, after a period of time, finds that 98 of them had a repeat criminal offense. Which
of the following represents the approximately probability that the recidivism rate at this prison is less than the
national average?

0.371 − 0.327 0.327 − 0.371

(A) 𝑃 𝑧> (B) 𝑃 𝑧<
√0.371(1 − 0.371) √0.327(1 − 0.327)
( 300 ) ( 300 )

0.327 − 0.371 0.371 − 0.327

(C) 𝑃 𝑧< (D) 𝑃 𝑧>
√0.371(1 − 0.371) √0.327(1 − 0.327)
( 300 ) ( 300 )

______ 6) In a test of 𝐻0 : 𝑝 = 0.35 with 𝛼 = 0.01, against 𝐻𝑎 : 𝑝 < 0.35, a large random sample produced a z-
score of −2.05. Based on this, which of the following conclusions can be drawn?

(A) We have convincing evidence that p < 0.35

(B) If the z-score was positive instead of negative, we would be able to reject the null hypothesis.
(C) We do not have convincing evidence that p = 0.35
(D) We cannot reject the null hypothesis in favor of the alternative

Page 21