DEPARTEMEN TEKNIK MESIN DAN INDUSTRI

STATISTIKA I

Tests of Hypotheses for a Single Sample

Erny Rahayu Wijayanti

erny.wijayanti@ugm.ac.id
Example : Bottled Water
Customer
Is there, on average, at least
660ml of water in each
bottle?

Volume of Water ≥ 𝟔𝟔𝟔𝟔𝟔𝟔 𝒎𝒎𝒎𝒎

Manufacturer
Is there, on average, exactly
660ml of water in each
bottle?
Volume of Water = 𝟔𝟔𝟔𝟔𝟔𝟔 𝒎𝒎𝒎𝒎

Collect 50 bottles from all

over province as SAMPLE
660 ml
TEST THE ASSUMPTION
Example : Hybrid Engine

Company Claim :

𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭 𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬𝑬 > 𝟑𝟑𝟑𝟑 𝒎𝒎𝒎𝒎𝒎𝒎

The manufacturer is
making a claim it wishes

An auto manufacturer has developed a new hybrid

to test, it is not testing an
engine technology, it claims reduces fuel assumption that already
consumption while driving in the city. The claim is
that the new technology improves fuel old engine exists.
that produces 30 mpg. The company will run
controlled tests to look for statistical evidence to
support the claim that the new engine offers Notice : Assumption vs Claim
better efficiency than the old model.
A. Hypothesis Testing
Statistical Hypothesis
A statistical hypothesis is a statement about the parameters of one or more
populations.

• By definition, the null and alternative hypotheses are opposite .

(The null is either rejected or it is not. Only if the null is rejected can we
proceed to the alternative).
• Researchers can start with either the null or the alternative and then form the
other as a complement to the first.
• Which to start with largely depends on the point of view of the researcher, the
context of the problem, and what can or cannot be assumed to be known
upfront.
Null and Alternative

Null Hypothesis Alternative Hypothesis

𝑯𝑯𝟎𝟎 𝑯𝑯𝒂𝒂 or 𝑯𝑯𝟏𝟏
“Null” means nothing new or The “alternative” is simply the
different with the assumption “other option” when the null is
maintained. rejected, nothing more.

Assumption, nothing new Rejection of an assumption

Assumed to be “true”; a given Rejection of an assumption or the given

Negation of the research question Research question to be “proven”
Always contains an equality (=, ≤, ≥) Does not contain equality (≠, <, >)
The alternative hypothesis is that one which the researcher wishes to support, while the
null hypothesis, a contradiction to the alternative hypothesis.
Null and Alternative
 All statistical conclusions are made in reference to the null hypothesis.
 As researchers we either reject the null hypothesis or fail to reject the
null hypothesis (we do not accept the null). This is due to the fact that
the null hypothesis is assumed to be true from the start, rejecting or
failing to reject an assumption.
 If we reject the null hypothesis, then we conclude the data supports the
alternative hypothesis.
 However if we fail to reject the null hypothesis, it does not mean we
have proven the null hypothesis is “true”
• Why? Because remember from the outset we assumed it was true
• Failure to reject the null does not equate to “proof” about its truth
Example : Bottled Water
A bottled water manufacturer
states on the product label that
each bottle contains 660 ml of
water. You work for a
government agency that
protects consumers by testing
product volumes. A sample of
50 bottles is tested.

THE ASSUMPTION
The 660 ml on the bottle is
If the data indicates the bottles are being filled assumed to be true.
properly, then we failed to reject the null
(failed to reject our assumption). HYPOTHESIS
𝑯𝑯𝟎𝟎 ∶ 𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽 = 𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔
We are not saying we have proven the null. 𝑯𝑯𝟏𝟏 ∶ 𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽 ≠ 𝟔𝟔𝟔𝟔𝟔𝟔 𝒎𝒎𝒎𝒎
Just that our assumption has held up.
Example : Bottled Water
If the data indicates the bottles are not being filled
properly, then we reject the null
(reject our assumption).
Our assumption has not held up under analysis. We
have statistical support for the validity of the
alternative hypothesis.
THE ASSUMPTION
The 660 ml on the bottle is
assumed to be true.
HYPOTHESIS
𝑯𝑯𝟎𝟎 ∶ 𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽 = 𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔𝟔
𝑯𝑯𝟏𝟏 ∶ 𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽𝑽 ≠ 𝟔𝟔𝟔𝟔𝟔𝟔 𝒎𝒎𝒎𝒎
If the data indicates the bottles are
being filled properly, then we
failed to reject the null
(failed to reject our assumption).
We are not saying we have proven
the null.
Just that our assumption has held
up.
Example : Down on the Farm
THE ASSUMPTION
There has been no change in
farm size since 2006.

We wish to see the farm size

has increased since 2006

According to the US Dept. of Agriculture, in 2006 HYPOTHESIS

the average farm size in the state of Texas was
2.3 𝑘𝑘𝑚𝑚2 . Since the decades-long trend as been for 𝑯𝑯𝟎𝟎 ∶ 𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺 = 𝟐𝟐. 𝟑𝟑 𝒌𝒌𝒎𝒎𝟐𝟐
farm sizes to increase due to large agribusiness, a 𝑯𝑯𝟏𝟏 ∶ 𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺 > 𝟐𝟐. 𝟑𝟑 𝒌𝒌𝒎𝒎𝟐𝟐
business analyst wishes to test if current (2023)
farm size is larger that it was in 2006.
Establish a null and alternative hypothesis!
Example : The Fire Alarm Hypothesis

Let’s say you are laying on your bed while scrolling your smart phone.
Suddenly you encounter a sudden smell of smoke.
You know that may mean a serious fire is taking place. Or it could be nothing
serious; maybe someone burned trash.

What do you do next?

 If you think the smoky smell is

nothing serious, you may decide your
assumption that everything is normal
is correct and you will not pull the fire
alarm.
 If you think the smoky smell is due to
a serious fire, you may reject your
assumption that everything is normal
and you will pull the fire alarm.
A. Hypothesis Testing
What may go wrong?

1. You smell smoke.

2. You think, “This is not normal” (reject
the assumption that everything is Type I Error
OK/ reject the null hypothesis).
3. Therefore you pull the fire alarm. Rejecting the null hypothesis 𝐻𝐻0
4. The building is evacuated and the fire
department arrives to investigate. when it is true is defined as a
5. After the investigation it is
determined there was no fire. You type I error.
“falsely” pulled the fire alarm.
6. When you rejected your assumption Incorrectly rejecting the null
that everything was OK, when it really hypothesis.
was OK, you committed Type I Error.
A “false alarm.”.
A. Hypothesis Testing
What else may go wrong?

1. You smell smoke.

2. You think, “It is probably someone
who burned their trash. No big deal”. Type II Error
3. Therefore you do not reject your
assumption (null hypothesis) that Failing to reject the null
everything is OK. You uphold the null.
4. But there is indeed a fire, no one is hypothesis 𝐻𝐻0 when it is false is
injured, but the building burn to the
ground. defined as a type II error.
5. When you failed to reject your
assumption that everything was OK, Incorrectly not rejecting the null
when it really was NOT OK, you hypothesis.
committed Type II Error. A “false
alarm.”.
The Fire Alarm Hypothesis
𝑯𝑯𝟎𝟎 = 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔 𝒊𝒊𝒊𝒊 𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂𝒂 𝒃𝒃𝒃𝒃𝒃𝒃 𝒏𝒏𝒏𝒏𝒏𝒏 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔, 𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆 𝒊𝒊𝒊𝒊 𝑶𝑶𝑶𝑶 𝒂𝒂𝒂𝒂 𝒖𝒖𝒖𝒖𝒖𝒖𝒖𝒖𝒖𝒖; 𝒏𝒏𝒏𝒏 𝒇𝒇𝒇𝒇𝒇𝒇𝒇𝒇
𝑯𝑯𝟏𝟏 = 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔 𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆 𝒐𝒐𝒐𝒐 𝒂𝒂 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔 𝒇𝒇𝒇𝒇𝒇𝒇𝒇𝒇, 𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆 𝑵𝑵𝑵𝑵𝑵𝑵 𝑶𝑶𝑶𝑶 𝒂𝒂𝒂𝒂 𝒖𝒖𝒖𝒖𝒖𝒖𝒖𝒖𝒖𝒖; 𝒇𝒇𝒇𝒇𝒇𝒇𝒇𝒇

Actual Condition
𝑵𝑵𝑵𝑵 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔 𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭 𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺 𝑭𝑭𝑭𝑭𝑭𝑭𝑭𝑭
Conclusion

Do not reject 𝐻𝐻0 Correct Conclusion Type II error

(no serious fire)
Reject 𝐻𝐻0 Type I error Correct Conclusion
(serious fire)

*In general, the real-world consequences of a Type II Error are much greater.
In this case, Type II error may mean loss of property or even lives.
A. Hypothesis Testing
Probability of Type I Error
𝛼𝛼 = 𝑃𝑃 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝐼𝐼 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 = 𝑃𝑃(𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝐻𝐻0 𝑤𝑤𝑤𝑤𝑤𝑤𝑤 𝐻𝐻0 𝑖𝑖𝑖𝑖 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡)

The type I error probability is called the significance level, the α-error
or size of the test.

Probability of Type II Error

𝛽𝛽 = 𝑃𝑃 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 𝐼𝐼𝐼𝐼 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 = 𝑃𝑃(𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑡𝑡𝑡𝑡 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟 𝐻𝐻0 𝑤𝑤𝑤𝑤𝑤𝑤𝑤 𝐻𝐻0 𝑖𝑖𝑖𝑖 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓)

The type II error probability is called the 𝜷𝜷 − 𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆

A. Hypothesis Testing

Decisions in Hypothesis Testing

Decision 𝑯𝑯𝟎𝟎 is TRUE 𝑯𝑯𝟎𝟎 is FALSE

Fail to reject 𝐻𝐻0 No Error Type II Error

Reject 𝐻𝐻0 Type I Error No Error

An Ejection Seat and a Rocket Motor
Suppose that an engineer is designing an air
crew escape system that consists of an
ejection seat and a rocket motor that powers
the seat. The rocket motor contains a
propellant, and for the ejection seat to
function properly, the propellant should have
a mean burning rate of 50 𝑐𝑐𝑐𝑐/sec. If the
burning rate is too low, the ejection seat may
not function properly, leading to an unsafe
ejection and possible injury of the pilot.
Higher burning rates may imply instability in
the propellant or an ejection seat that is too
powerful, again leading to possible pilot
injury.
Does the mean burning rate of the
propellant equal 𝟓𝟓𝟓𝟓 𝒄𝒄𝒄𝒄/𝒔𝒔𝒔𝒔𝒔𝒔, or is it some
other value (either higher or lower)?
An Ejection Seat and a Rocket Motor
𝑯𝑯𝟎𝟎 : 𝝁𝝁 = 𝟓𝟓𝟓𝟓 𝒄𝒄𝒄𝒄 𝒑𝒑𝒑𝒑𝒑𝒑 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔
𝑯𝑯𝟏𝟏 : 𝝁𝝁 ≠ 𝟓𝟓𝟓𝟓 𝒄𝒄𝒄𝒄 𝒑𝒑𝒑𝒑𝒑𝒑 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔

Suppose :
Size of sample 𝑛𝑛 = 10 𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠𝑠
Standard deviation of the burning rate
𝜎𝜎 = 2.5 𝑐𝑐𝑐𝑐/𝑠𝑠𝑠𝑠𝑠𝑠
The sample mean burning rate 𝑥𝑥̅ is observed
If 48.5 ≤ 𝑥𝑥̅ ≤ 51.5, we will not reject the null
hypothesis.
An Ejection Seat and a Rocket Motor

The probability of making a type I error :

� < 𝟒𝟒𝟒𝟒. 𝟓𝟓 𝒘𝒘𝒘𝒘𝒘𝒘𝒘𝒘 𝝁𝝁 = 𝟓𝟓𝟓𝟓 + 𝑷𝑷 𝑿𝑿
𝜶𝜶 = 𝑷𝑷 𝑿𝑿 � > 𝟓𝟓𝟓𝟓. 𝟓𝟓 𝒘𝒘𝒘𝒘𝒘𝒘𝒘𝒘 𝝁𝝁 = 𝟓𝟓𝟓𝟓

The z-values that correspond to the critical values 48.5 and 51.5 are

𝟒𝟒𝟒𝟒. 𝟓𝟓 − 𝟓𝟓𝟓𝟓 𝟓𝟓𝟓𝟓. 𝟓𝟓 − 𝟓𝟓𝟓𝟓

𝒁𝒁𝟏𝟏 = = −𝟏𝟏. 𝟗𝟗𝟗𝟗 𝒁𝒁𝟐𝟐 = = 𝟏𝟏. 𝟗𝟗𝟗𝟗
𝟐𝟐. 𝟓𝟓/√𝟏𝟏𝟏𝟏 𝟐𝟐. 𝟓𝟓/√𝟏𝟏𝟏𝟏

Therefore,
𝜶𝜶 = 𝑷𝑷 𝒛𝒛 < −𝟏𝟏. 𝟗𝟗𝟗𝟗 + 𝑷𝑷 𝒛𝒛 > 𝟏𝟏. 𝟗𝟗𝟗𝟗 = 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 + 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 = 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎
An Ejection Seat and a Rocket Motor

The critical region for 𝑯𝑯𝟎𝟎 : 𝝁𝝁 = 𝟓𝟓𝟓𝟓 𝑽𝑽𝑽𝑽 𝑯𝑯𝟏𝟏 : 𝝁𝝁 ≠ 𝟓𝟓𝟓𝟓

This implies that 5.74% of all random samples would lead to rejection of
the hypothesis 𝐻𝐻0 : 𝜇𝜇 = 50𝑐𝑐𝑐𝑐/𝑠𝑠𝑠𝑠𝑠𝑠 when the true mean burning rate is really
50cm/sec.
An Ejection Seat and a Rocket Motor

Acceptance region

The critical region for 𝑯𝑯𝟎𝟎 : 𝝁𝝁 = 𝟓𝟓𝟓𝟓 𝑽𝑽𝑽𝑽 𝑯𝑯𝟏𝟏 : 𝝁𝝁 ≠ 𝟓𝟓𝟓𝟓

Question!
1. What will happen if 𝑛𝑛 = 16, 𝑛𝑛 = 20, 𝑛𝑛 = 50?
2. What will happen if we are widening the acceptance region?
An Ejection Seat and a Rocket Motor

The probability of making a type II error :

𝛽𝛽 = 𝑃𝑃 48.5 ≤ 𝑋𝑋� ≤ 51.5 𝑤𝑤𝑤𝑤𝑤𝑤𝑤 𝜇𝜇 = 52

The z-values that correspond to the critical values 48.5 and 51.5 when
𝜇𝜇 = 52 are
48.5 − 52 51.5 − 52
𝑍𝑍1 = = −4.43 𝑍𝑍2 = = −0.63
2.5/√10 2.5/√10

Therefore,
𝜷𝜷 = 𝑷𝑷 −𝟒𝟒. 𝟒𝟒𝟒𝟒 ≤ 𝒁𝒁 ≤ −𝟎𝟎. 𝟔𝟔𝟔𝟔 = 𝑷𝑷 𝒁𝒁 ≤ −𝟎𝟎. 𝟔𝟔𝟔𝟔 − 𝑷𝑷 𝒁𝒁 ≤ −𝟒𝟒. 𝟒𝟒𝟒𝟒
= 𝟎𝟎. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐 − 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎𝟎𝟎 = 𝟎𝟎. 𝟐𝟐𝟐𝟐𝟐𝟐𝟐𝟐
An Ejection Seat and a Rocket Motor

Question!
What will happen to 𝜷𝜷 − 𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆𝒆 if 𝒏𝒏 = 𝟏𝟏𝟏𝟏, 𝒏𝒏 = 𝟐𝟐𝟐𝟐, 𝒏𝒏 = 𝟓𝟓𝟓𝟓?
A. Hypothesis Testing
Important Points :
1. The size of the critical region, and consequently the probability of
a type I error 𝛼𝛼, can always be reduced by appropriate selection
of the critical values.
2. Type I and type II errors are related. A decrease in the probability
of one type of error always results in an increase in the
probability of the other provided that the sample size 𝑛𝑛 does not
change.
3. An increase in sample size reduces 𝛽𝛽 provided that 𝛼𝛼 is held
constant.
4. When the null hypothesis is false, 𝛽𝛽 increases as the true value of
the parameter approaches the value hypothesized in the null
hypothesis. The value of 𝛽𝛽 decreases as the difference between
the true mean and the hypothesized value increases
A. Hypothesis Testing
A widely used procedure in hypothesis testing is to use a type I error or
significance level of 𝛼𝛼 = 0.05 . This value has evolved through
experience and may not be appropriate for all situations.

Power
The power of a statistical test is the probability of rejecting
the null hypothesis 𝐻𝐻0 when the alternative hypothesis is
true.
𝑷𝑷𝑷𝑷𝑷𝑷𝑷𝑷𝑷𝑷 = 𝟏𝟏 − 𝜷𝜷
(the probability of correctly rejecting a false null hypothesis)
𝑃𝑃-Value
The 𝑷𝑷-value is the smallest level of significance that would
lead to rejection of the null hypothesis 𝐻𝐻0 with the given
data.
𝑃𝑃-Value (Observed Significance level)
Consider the two-sided hypothesis test for burning rate
𝑯𝑯𝟎𝟎 : 𝝁𝝁 = 𝟓𝟓𝟓𝟓 𝒄𝒄𝒄𝒄 𝒑𝒑𝒑𝒑𝒑𝒑 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔
𝑯𝑯𝟏𝟏 : 𝝁𝝁 ≠ 𝟓𝟓𝟓𝟓 𝒄𝒄𝒄𝒄 𝒑𝒑𝒑𝒑𝒑𝒑 𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔𝒔
with 𝑛𝑛 = 16 and 𝜎𝜎 = 2.5 𝑐𝑐𝑐𝑐/𝑠𝑠𝑠𝑠𝑠𝑠. Suppose that the observed sample mean is
𝑥𝑥̅ = 51.3 𝑐𝑐𝑐𝑐/𝑠𝑠𝑠𝑠𝑠𝑠.

P-value is the area of the shaded region when 𝑥𝑥̅ = 5103

� < 𝟓𝟓𝟓𝟓. 𝟑𝟑
𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯 = 𝟏𝟏 − 𝑷𝑷 𝟒𝟒𝟒𝟒. 𝟕𝟕 < 𝑿𝑿

𝟒𝟒𝟒𝟒. 𝟕𝟕 − 𝟓𝟓𝟓𝟓 𝟓𝟓𝟓𝟓. 𝟑𝟑 − 𝟓𝟓𝟓𝟓

= 𝟏𝟏 − 𝑷𝑷 < 𝒁𝒁 <
𝟐𝟐. 𝟓𝟓/√𝟏𝟏𝟏𝟏 𝟐𝟐. 𝟓𝟓/√𝟏𝟏𝟏𝟏
= 𝟏𝟏 − 𝑷𝑷 −𝟐𝟐. 𝟎𝟎𝟎𝟎 < 𝒁𝒁 < 𝟐𝟐. 𝟎𝟎𝟎𝟎
= 𝟏𝟏 − 𝟎𝟎. 𝟗𝟗𝟗𝟗𝟗𝟗 = 𝟎𝟎. 𝟎𝟎𝟎𝟎𝟎𝟎
B. Tests on the Mean of a Normal Distribution, Variance Known
Testing Hypotheses on the Mean, Variance Known (Z-Tests)
 Test Hypothesis  Test Hypothesis  Test Hypothesis
𝐻𝐻0 : 𝜇𝜇 = 𝜇𝜇0 𝐻𝐻0 : 𝜇𝜇 ≤ 𝜇𝜇0 𝐻𝐻0 : 𝜇𝜇 ≥ 𝜇𝜇0
𝐻𝐻1 : 𝜇𝜇 ≠ 𝜇𝜇0 𝐻𝐻1 : 𝜇𝜇 > 𝜇𝜇0 𝐻𝐻1 : 𝜇𝜇 < 𝜇𝜇0

 Test Statistic :
� −𝝁𝝁𝟎𝟎
𝑿𝑿
𝒁𝒁𝟎𝟎 =
 𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯 𝝈𝝈/ 𝒏𝒏
Probability above 𝑧𝑧0 and
 𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯  𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯
probability below − 𝑧𝑧0 ,
Probability above 𝑧𝑧0 Probability above 𝑧𝑧0
𝑃𝑃 = 2[1 − Φ 𝑧𝑧0 ]
𝑃𝑃 = 1 − Φ 𝑧𝑧0 𝑃𝑃 = Φ 𝑧𝑧0
 Rejection Criterion for Fixed-
 Rejection Criterion for  Rejection Criterion for
Level Tests
Fixed-Level Tests Fixed-Level Tests
𝑧𝑧0 > 𝑧𝑧𝛼𝛼 or 𝑧𝑧0 < −𝑧𝑧𝛼𝛼 𝑧𝑧0 > 𝑧𝑧𝛼𝛼 𝑧𝑧0 < −𝑧𝑧𝛼𝛼
2 2
Example : Propellant Burning Rate

Air crew escape systems are powered by a solid propellant. The

burning rate of this propellant is an important product
characteristic. Specifications require that the mean burning rate
must be 50 𝑐𝑐𝑐𝑐/𝑠𝑠𝑠𝑠𝑠𝑠. We know that the standard deviation of
burning rate is 𝜎𝜎 = 2 𝑐𝑐𝑐𝑐/𝑠𝑠𝑠𝑠𝑠𝑠.
The experimenter decides to specify a type I error probability or
significance level of 𝛼𝛼 = 0.05 and selects a random sample of
𝑛𝑛 = 25 and obtains a sample average burning rate of x� =
51.3 𝑐𝑐𝑐𝑐/𝑠𝑠𝑠𝑠𝑠𝑠 .
What conclusions should be drawn?
Example : Propellant Burning Rate
 Test Hypothesis
𝐻𝐻0 : 𝜇𝜇 = 50
𝐻𝐻1 : 𝜇𝜇 ≠ 50
 Significance level 𝛼𝛼 = 0.05
 Test Statistic
� − 𝝁𝝁𝟎𝟎 𝟓𝟓𝟓𝟓. 𝟑𝟑 − 𝟓𝟓𝟓𝟓
𝑿𝑿
𝒛𝒛𝟎𝟎 = = = 𝟑𝟑. 𝟐𝟐𝟐𝟐
𝝈𝝈/ 𝒏𝒏 𝟐𝟐/ 𝟐𝟐𝟐𝟐
 Rejection Criterion for Fixed-Level Tests

𝑧𝑧0 > 𝑧𝑧𝛼𝛼 or 𝑧𝑧0 < −𝑧𝑧𝛼𝛼 , where 𝑧𝑧𝛼𝛼 = 𝑧𝑧0.025 = 1.96
2 2 2
 𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯
𝑃𝑃 = 2 1 − Φ 𝑧𝑧0 = 2 1 − Φ 3.25 = 0.0012

Because of 𝑧𝑧0 = 3.25 > 𝑧𝑧0.025 = 1.96, then we reject 𝐻𝐻0 = 50 at the 0.05
level of significance.
Example : Propellant Burning Rate

Interpretation:
We conclude that the mean burning rate differs from 50
centimeters per second, based on a sample of 25
measurements. In fact, there is strong evidence that the
mean burning rate is not 50 centimeters per second.
Sample Size
Probability of a Type II Error for a Two-Sided Test on the Mean, Variance Known

𝜇𝜇 − 𝜇𝜇0 𝑛𝑛 𝜇𝜇 − 𝜇𝜇0 𝑛𝑛
𝛽𝛽 = Φ 𝑧𝑧𝛼𝛼/2 − − Φ −𝑧𝑧𝛼𝛼/2 −
𝜎𝜎 𝜎𝜎

Sample Size for a Two-Sided Test on the Mean, Variance Known

𝑧𝑧𝛼𝛼 + 𝑧𝑧𝛽𝛽 2 𝜎𝜎 2
2
𝑛𝑛 ≈ 2
𝜇𝜇 − 𝜇𝜇0

Sample Size for a One-Sided Test on the Mean, Variance Known

2
𝑧𝑧𝛼𝛼 + 𝑧𝑧𝛽𝛽 𝜎𝜎 2
𝑛𝑛 ≈
𝜇𝜇 − 𝜇𝜇0 2
C. Tests on the Mean of a Normal Distribution, Variance Unknown

Testing Hypotheses on the Mean, Variance Unknown (T-Tests)

 Test Hypothesis  Test Hypothesis  Test Hypothesis
𝐻𝐻0 : 𝜇𝜇 = 𝜇𝜇0 𝐻𝐻0 : 𝜇𝜇 ≤ 𝜇𝜇0 𝐻𝐻0 : 𝜇𝜇 ≥ 𝜇𝜇0
𝐻𝐻1 : 𝜇𝜇 ≠ 𝜇𝜇0 𝐻𝐻1 : 𝜇𝜇 > 𝜇𝜇0 𝐻𝐻1 : 𝜇𝜇 < 𝜇𝜇0

 Test Statistic
� − 𝝁𝝁𝟎𝟎
𝑿𝑿
𝒕𝒕𝟎𝟎 =
𝑺𝑺/ 𝒏𝒏
 𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯
Probability above 𝑡𝑡0 and
 𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯  𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯
probability below − 𝑡𝑡0 ,
Probability above 𝑧𝑧0 Probability above 𝑧𝑧0
 Rejection Criterion for Fixed-
 Rejection Criterion for  Rejection Criterion for
Level Tests
Fixed-Level Tests Fixed-Level Tests
𝑡𝑡0 > 𝑡𝑡𝛼𝛼,𝑛𝑛−1 or 𝑡𝑡0 < −𝑡𝑡𝛼𝛼,𝑛𝑛−1 𝑡𝑡0 > 𝑡𝑡𝛼𝛼,𝑛𝑛−1 𝑡𝑡0 > 𝑡𝑡𝛼𝛼,𝑛𝑛−1
2 2
Example : Golf Club Design
The increased availability of light materials with high strength has
revolutionized the design and manufacture of golf clubs, particularly
drivers. Clubs with hollow heads and very thin faces can result in much
longer tee shots, especially for players of modest skills. This is due
partly to the “spring-like effect” that the thin face imparts to the ball.
Firing a golf ball at the head of the club and measuring the ratio of the
ball’s outgoing velocity to the incoming velocity can quantify this
spring-like effect. The ratio of velocities is called the coefficient of
restitution of the club. An experiment was performed in which 15
drivers produced by a particular club maker were selected at random
and their coefficients of restitution measured. In the experiment, the
golf balls were fired from an air cannon so that the incoming velocity
and spin rate of the ball could be precisely controlled. It is of interest
to determine whether there is evidence (with 𝛼𝛼 = 0.05) to support a
claim that the mean coefficient of restitution exceeds 0.82. The
observations follow:
Example : Golf Club Design

The observations follow:

0.8411 0.8191 0.8182 0.8125 0.8750
0.8580 0.8532 0.8483 0.8276 0.7983
0.8042 0.8730 0.8282 0.8359 0.8660
Example : Golf Club Design
Example : Golf Club Design
Example : Golf Club Design

Interpretation :
There is strong evidence
to conclude that the mean
coefficient of restitution
exceeds 0.82.
E. Tests on a Population Proportion

Testing Hypotheses on a Binomial Proportion

 Test Hypothesis  Test Hypothesis  Test Hypothesis
𝐻𝐻0 : 𝑝𝑝 = 𝑝𝑝0 𝐻𝐻0 : 𝑝𝑝 ≤ 𝑝𝑝0 𝐻𝐻0 : 𝑝𝑝 ≥ 𝑝𝑝0
𝐻𝐻1 : 𝑝𝑝 ≠ 𝑝𝑝0 𝐻𝐻1 : 𝑝𝑝 > 𝑝𝑝0 𝐻𝐻1 : 𝑝𝑝 < 𝑝𝑝0

 Test Statistic :
𝑿𝑿 − 𝒏𝒏𝒑𝒑𝟎𝟎
𝒁𝒁𝟎𝟎 =
𝒏𝒏𝒑𝒑𝟎𝟎 (𝟏𝟏 − 𝒑𝒑𝟎𝟎 )
 𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯
Probability above 𝑧𝑧0 and
 𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯  𝑷𝑷 − 𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯𝐯
probability below − 𝑧𝑧0 ,
Probability above 𝑧𝑧0 Probability above 𝑧𝑧0
𝑃𝑃 = 2[1 − Φ 𝑧𝑧0 ]
𝑃𝑃 = 1 − Φ 𝑧𝑧0 𝑃𝑃 = Φ 𝑧𝑧0
 Rejection Criterion for Fixed-
 Rejection Criterion for  Rejection Criterion for
Level Tests
Fixed-Level Tests Fixed-Level Tests
𝑧𝑧0 > 𝑧𝑧𝛼𝛼 or 𝑧𝑧0 < −𝑧𝑧𝛼𝛼
2 2 𝑧𝑧0 > 𝑧𝑧𝛼𝛼 𝑧𝑧0 < −𝑧𝑧𝛼𝛼
Example :
Example : Automobile Engine Controller
Exercise 1
Exercise 3
Assignment 1
Assignment 2
Cloud seeding has been studied for many decades as a weather modification procedure
(for an interesting study of this subject, see the article in Technometrics, “A Bayesian
Analysis of a Multiplicative Treatment Effect in Weather Modification,” 1975, Vol. 17, pp. 161–
166). The rainfall in acre-feet from 20 clouds that were selected at random and seeded
with silver nitrate follows:

18.0, 30.7, 19.8, 27.1, 22.3, 18.8, 31.8, 23.4, 21.2, 27.9, 31.9, 27.1, 25.0, 24.7, 26.9, 21.8, 29.2,
34.8, 26.7, and 31.6.

a. Can you support a claim that mean rainfall from seeded clouds exceeds 25 acre-
feet? Use α = 0.01. Find the P-value.
b. Check that rainfall is normally distributed.
c. Compute the power of the test if the true mean rainfall is 27 acre-feet.
d. What sample size would be required to detect a true mean rainfall of 27.5 acre-feet
if you wanted the power of the test to be at least 0.9?
e. Explain how the question in part (a) could be answered by constructing a one-sided
confidence bound on the mean diameter.
Assignment 3
It is believed that at least 60% of the residents in a certain
area favor an annexation suit by a neighboring city. What
conclusion would you draw if only 110 in a sample of 200
voters favored the suit? Use a 0.05 level of significance.