Test of Hypotheses: Hypothesis

Kathmandu University
Course: MATH 208, ENVE II/II
Prepared by: Dr. Samir Shrestha
Test of Hypotheses
Hypothesis
A hypothesis is an educated guess about something in the world around you. It should be
testable, either by experiment or observation. For example:
1. A new medicine you think might work better than existing one.
2. A new way of teaching technique you think might be better.
Hypothesis Testing
A hypothesis testing is a statistical test that is used to determine whether there is enough
evidence in a sample of data to infer that a certain condition is true for the entire population.
Or
Hypothesis testing is the use of statistics to determine the probability that a given hypothesis is
true
The general idea of hypothesis testing

 Making an initial assumption.
 Collecting evidence(s) (data).
 Based on the available evidence (data), deciding whether to reject or not reject the initial
assumption.
Example: Is normal body temperature really 98.6 degrees F?
A researcher hypothesized that the average adult body temperature is not equal to the claimed
98.6 degrees F. That is, the researcher wants an answer to the question: "Is the average adult
body temperature 98.6 degrees? Or is it different?" To answer his research question, the
researcher starts by assuming that the average adult body temperature was 98.6 degrees F.
(Initial Assumption)
Then, the researcher went out and tried to find evidence that refutes his initial assumption. In
doing so, he selects a random sample of 130 adults. The average body temperature of the 130
sampled adults is 98.25 degrees. (Collecting the Evidence)
Then, the researcher uses the data he collected to make a decision about his initial assumption.
There are:
Two Possibilities in Decision:
By Analyzing the evidence (observed data)

1
 Either, He reject the initial Assumption
 Or, He does not reject the initial assumption (or he says there is no enough evidence to
reject the initial assumption).
Statement of Hypotheses
Initial Assumption is known as Null Hypothesis and denoted by 𝑯𝟎 and alternate to (opposite
to) to Null Hypothesis is known as Alternate Hypothesis and denoted by 𝑯𝟏 .
In this research question: (Constructing the Hypothesis)
𝑁𝑢𝑙𝑙 𝐻𝑦𝑝𝑜𝑡𝑕𝑒𝑠𝑖𝑠 𝐻0 : 𝜇 = 98.6
𝐴𝑙𝑡𝑒𝑟𝑛𝑎𝑡𝑒 𝐻𝑦𝑝𝑜𝑡𝑕𝑒𝑠𝑖𝑠 𝐻1 : 𝜇 ≠ 98.6
Example: Criminal Trial

Criminal justice system assumes "the defendant (suspect) is innocent until proven guilty." That
is, our initial assumption is that the defendant is innocent.
In the practice of statistics, we make our initial assumption when we state our two competing
hypotheses - the null hypothesis (𝑯𝟎 ) and the alternative hypothesis (𝑯𝟏 ). Here, our hypotheses
are:
𝑯𝟎 : Defendant is not guilty (innocent)
𝑯𝟏 : Defendant is guilty
In statistics, we always assume the null hypothesis is true. That is, the null hypothesis is always
our initial assumption.
The prosecution team then collects evidence — such as finger prints, blood spots, hair samples,
carpet fibers, shoe prints, ransom notes, and handwriting samples — with the hopes of finding
"sufficient evidence" to make the assumption of innocence refutable.
In statistics, the data are the evidence.
The jury then makes a decision based on the available evidence:
If the jury finds sufficient evidence — beyond a reasonable doubt — to make the assumption of
innocence refutable, the jury rejects the null hypothesis and deems the defendant guilty. We
behave as if the defendant is guilty.
If there is insufficient evidence, then the jury does not reject the null hypothesis. We behave as if
the defendant is innocent.
In statistics, we always make one of two decisions. We either "reject the null hypothesis" or
we "fail to reject the null hypothesis."
2
Errors in Hypothesis Testing
We make our decision based on evidence not on 100% guaranteed proof.
That means,
If we reject the null hypothesis, we do not prove that the alternative hypothesis is true.
If we do not reject the null hypothesis, we do not prove that the null hypothesis is true.
We merely state that there is enough evidence to behave one way or the other. This is always
true in statistics! Because of this, whatever the decision, there is always a chance that we made
an error.
Let's review the two types of errors that can be made in criminal trials:
Truth
Jury Decision Not Guilty Guilty

Not Guilty OK ERROR (Type II)
Guilty ERROR (Type I) OK
and let's see how they correspond to the two types of errors in hypothesis testing:
Truth
Decision Null Hypothesis Alternative Hypothesis

Do not reject Null Hypothesis No Error Type II - ERROR
(𝟏 − 𝜶) (𝜷)
Reject Null Hypothesis Type I - ERROR No Error
(𝜶) (𝟏 − 𝜷)
Note that, in statistics, we call the two types of errors by two different names: one is called a "Type I
error," and the other is called a "Type II error." Here are the formal definitions of the two types of errors:
Type I error: The null hypothesis is rejected when it is true.
Type II error: The null hypothesis is not rejected when it is false.
𝑃 𝑇𝑦𝑝𝑒 𝐼 𝑒𝑟𝑟𝑜𝑟 = 𝑃(𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑤𝑕𝑒𝑛 𝐻0 𝑖𝑠 𝑡𝑟𝑢𝑒 ) = 𝛼 (Called Level of significance)
𝑃 𝑇𝑦𝑝𝑒 𝐼𝐼 𝑒𝑟𝑟𝑜𝑟 = 𝑃(𝐹𝑎𝑖𝑙𝑒𝑑 𝑡𝑜 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑤𝑕𝑒𝑛 𝐻0 𝑖𝑠 𝑓𝑎𝑙𝑠𝑒 ) = 𝛽
Definition: (Power)
The power of a statistical test is the probability of rejecting the null hypothesis 𝐻0 when the
alternative hypothesis is true. The power is computed as 1 − 𝛽, and power can be interpreted as
the probability of correctly rejecting a false null hypothesis.
3
That means, 𝑃𝑜𝑤𝑒𝑟 = 1 − 𝛽 = 𝑃(𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑤𝑕𝑒𝑛 𝐻0 𝑖𝑠 𝑓𝑎𝑙𝑠𝑒 )
Making the decision (Two Approaches)

In statistics, there are two ways to determine to reject the Null Hypothesis or not to reject the
Null Hypothesis:
 Critical value approach (Used in many of the older textbooks). or

 P-value approach (what is used most often in research, journal articles, and statistical
software).
Decision can be made by either of the approaches by calculating the appropriate test statistics.
There are basically four steps to do Hypothesis testing:

Step 1: State the Null hypothesis 𝐻0 and Alternate Hypothesis 𝐻1
Step 2: State the 𝛼 level of significance (mostly given to be 𝛼 = 0.1 𝑜𝑟 0.05 𝑜𝑟 0.01). If you aren’t given
a 𝛼 level of significance, use 5% (𝛼 = 0.05). (Calculate critical value corresponding to 𝛼 if using critical
value approach to make decision)
Step 3: Compute the test statistic (For example 𝑧 − 𝑣𝑎𝑙𝑢𝑒 𝑜𝑟 𝑡 − 𝑣𝑎𝑙𝑢𝑒 𝑜𝑟 𝜒 2 − 𝑣𝑎𝑙𝑢𝑒 𝑜𝑟 𝐹 − 𝑣𝑎𝑙𝑢𝑒
depending upon the research questions using the appropriate formula) using the information provided
(data). (Calculate P-value is using P-value approach to make decision)
Step 4: Make decision (Either by Critical Value Approach or P-Value Approach)
Two-Sided and One-Sided Hypotheses Examples

Two-Sided Hypotheses (Two tail test)
Example1: Blood glucose levels for obese (over weight) patients have a mean of 100 with a
standard deviation of 15. A researcher thinks that a diet high in raw cornstarch will have a
positive or negative effect on blood glucose levels. Samples of 30 patients who have tried the
raw cornstarch diet have a mean glucose level of 140. State the Null and Alternative Hypotheses.
Answer
Null hypothesis: 𝐻0 : 𝜇 = 100
Alternate hypothesis: 𝐻1 : 𝜇 ≠ 100
Example 2: A manufacturing process produces ball bearings with diameters that have a normal
distribution with known standard deviation of 0.04 centimeters. Ball bearings with diameters that
are too small or too large are undesirable. In order to test the claim that μ = 0.50 centimeters.
Assume that a random sample of 25 gave a mean diameter of 0.51 centimeters. State Null and
Alternative hypotheses.
Answer:
4
Null hypothesis: 𝐻0 : 𝜇 = 0.50
Alternate hypothesis: 𝐻1 : 𝜇 ≠ 0.50
One-sided Hypothesis (right tail test)
Example1: A principal at a certain school claims that the students in his school are above
average intelligence. A random sample of thirty students IQ scores have a mean score of 112.
The mean population IQ is 100 with a standard deviation of 15. State Null and Alternative
Hypotheses.
Answer:
Null hypothesis: 𝐻0 : 𝜇 = 100 (or 𝜇 ≤ 100)
Alternate hypothesis: 𝐻1 : 𝜇 > 100
Example2: Doctors believe that the teen sleeps on average 10 hours per day. A researcher
believes that teen on average sleep longer. Write null and alternative hypothesis.
Answer:
Null hypothesis: 𝐻0 : 𝜇 = 10 (or 𝜇 ≤ 10)
Alternate hypothesis: 𝐻1 : 𝜇 > 10
One-sided Hypothesis (left tail test)
Example 1: A certain chemical pollutant in the Genesee River has been constant for several
years with mean μ = 34 ppm (parts per million) and standard deviation σ = 8 ppm. A group of
factory representatives whose companies discharge liquids into the river is now claiming that
they have lowered the average with improved filtration devices. A group of environmentalists
will test to see if this is true at the 5% level of significance. Assume that their sample of size 50
gives a mean of 32.5 ppm. Write Null and Alternative hypotheses.
Answer:
Null hypothesis: 𝐻0 : 𝜇 = 34 (or 𝜇 ≥ 34)
Alternate hypothesis: 𝐻1 : 𝜇 < 34
Single Sample Test

Test of Hypothesis on mean 𝝁 on a Single Sample with known variance 𝝈𝟐
1. Test of Hypothesis on the mean of a Normal Distribution of known Variance - Two tail
test
Let us consider the given population X normally distributed with unknown mean 𝜇 but given
variance 𝜎 2
5
We are interested to test the following Hypothesis:
Step1: (State the Null and Alternative hypotheses)
Null hypothesis: 𝐻0 : 𝜇 = 𝜇0
Alternate hypothesis: 𝐻1 : 𝜇 ≠ 𝜇0
where 𝜇0 is a specified constant.
Step2: (Provide the Level of significance 𝜶 )
Level of significance 𝛼 could be given if not set its value. Most used ones are 𝛼 = 0.1 𝑜𝑟 𝛼 =
0.05 𝑜𝑟 𝛼 = 0.01. (If you are using critical value approach for decision find 𝑧𝛼/2 using z-table)
Table for 𝒛𝜶/𝟐
𝜶 𝒛𝜶/𝟐
0.1 𝒛𝟎.𝟎𝟓 = 𝟏. 𝟔𝟒
0.05 𝒛𝟎.𝟎𝟐𝟓 = 𝟏. 𝟗𝟔
0.01 𝒛𝟎.𝟎𝟎𝟓 = 𝟐. 𝟓𝟖
Step3: (Test Statistics):
Let 𝑋1 , 𝑋2 , … , 𝑋𝑛 is random sample taken from the normal population and compute the sample
1 𝑛 𝑋 −𝜇
average 𝑋 = 𝑛 𝑖=1 𝑋𝑖 . For this case we compute the test statistic 𝑧0 = 𝜎/√𝑛 . (If you want to use
P-value approach, calculate the probability 𝑝 = 2 ∗ 𝑃 𝑍 ≥ 𝑧0 , 𝑢𝑠𝑖𝑛𝑔 𝑧 − 𝑡𝑎𝑏𝑙𝑒).
Step4: (Decision Making)
You can use either Critical Value Approach or P-value Approach to make the decision about
the hypothesis. Both will provide the same answer regarding the decision.
Critical Value Approach to make decision
 Reject the null hypothesis 𝐻0 if either 𝑧0 > 𝑧𝛼/2 or 𝑧0 < −𝑧𝛼/2 (𝑂𝑟 𝑧0 > 𝑧𝛼/2 )
 Fail to reject the null hypothesis 𝐻0 if −𝑧𝛼/2 ≤ 𝑧0 ≤ 𝑧𝛼/2
(Equivalent) P-Value Approach to make decision
 Reject the null hypothesis 𝐻0 if 𝑝 < 𝛼

 Fail to reject the null hypothesis 𝐻0 𝑝 ≥ 𝛼
2. Test of Hypothesis on the mean of a Normal Distribution of known Variance – One tail
test (Right tail)
6
Alternate hypothesis: 𝐻1 : 𝜇 > 𝜇0
Level of significance 𝛼 could be given if not set its value. Most used ones are𝛼 = 0.1 𝑜𝑟 𝛼 =
0.05 𝑜𝑟 𝛼 = 0.01. (If you are using critical value approach for decision compute 𝑧𝛼 𝑢𝑠𝑖𝑛𝑔 𝑧 −
𝑡𝑎𝑏𝑙𝑒).
Table for 𝒛𝜶
𝜶 𝒛𝜶
0.1 𝒛𝟎.𝟏 = 𝟏. 𝟐𝟗
0.05 𝒛𝟎.𝟎𝟓 = 𝟏. 𝟔𝟒
0.01 𝒛𝟎.𝟎𝟏 = 𝟐. 𝟑𝟑
1 𝑛 𝑋 −𝜇
average 𝑋 = 𝑛 𝑖=1 𝑋𝑖 . For this case we compute the test statistic 𝑧0 = 𝜎/√𝑛 . (If you want to use
P-value approach, calculate the probability 𝑝 = 𝑃 𝑍 ≥ 𝑧0 𝑢𝑠𝑖𝑛𝑔 𝑧 − 𝑡𝑎𝑏𝑙𝑒).
 Reject the null hypothesis 𝐻0 if either 𝑧0 > 𝑧𝛼

 Fail to reject the null hypothesis 𝐻0 if 𝑧0 ≤ 𝑧𝛼

3. Test of Hypothesis on the mean of a Normal Distribution of known Variance – One tail
test (Left tail)

7
Alternate hypothesis: 𝐻1 : 𝜇 < 𝜇0
0.05 𝑜𝑟 𝛼 = 0.01. (If you are using critical value approach for decision compute 𝑧𝛼 𝑢𝑠𝑖𝑛𝑔 𝑧 −
Table for 𝒛𝜶
𝜶 𝒛𝜶
0.1 𝒛𝟎.𝟏 = 𝟏. 𝟐𝟗
0.05 𝒛𝟎.𝟎𝟓 = 𝟏. 𝟔𝟒
0.01 𝒛𝟎.𝟎𝟏 = 𝟐. 𝟑𝟑
1 𝑛 𝑋 −𝜇
average 𝑋 = 𝑛 𝑖=1 𝑋𝑖 . For this case we compute the test statistic 𝑧0 = 𝜎/√𝑛
( 𝑧0 𝑤𝑖𝑙𝑙 𝑏𝑒 𝑛𝑒𝑔𝑎𝑡𝑖𝑣𝑒). (If you want to use P-value approach, calculate the probability 𝑝 =
𝑃 𝑍 ≤ 𝑧0 𝑢𝑠𝑖𝑛𝑔 𝑧 − 𝑡𝑎𝑏𝑙𝑒).
 Reject the null hypothesis 𝐻0 if either 𝑧0 < −𝑧𝛼

 Fail to reject the null hypothesis 𝐻0 if 𝑧0 ≥ −𝑧𝛼

8
Example 1
Aircrew 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
centimeters per second. We know that the standard deviation of burning rate is 𝜎 =
2 centimeters per second. The experimenter decides to specify a type I error probability or
significance level of 𝛼 = 0.05 and selects a random sample of 𝑛 = 25 and obtains a sample
average burning rate of 𝑥 = 51.3 centimeters per second. What conclusions should be drawn?
Solution:
Step 1: (State Null and Alternative Hypothesis)
Null Hypothesis, 𝐻0 : 𝜇 = 50
Alternative Hypothesis, 𝐻1 : 𝜇 ≠ 50
Level of significance 𝛼 = 0.05. It’s a two sided z-test, we need 𝑧𝛼/2 = 𝑧0.025 = 1.96
We are provided the population standard deviation 𝜎 = 2, sample mean 𝑥 = 51.3 and sample
size 𝑛 = 25, 𝜇0 = 50. The test statistics is
𝑋 − 𝜇0 51.3 − 50
𝑧0 = = = 3.25
𝜎/√𝑛 2/√25

Critical Value Approach
Since 𝑧0 = 3.25 > 1.96 = 𝑧𝛼 /2 , we reject the null hypothesis 𝐻0 : 𝜇 = 50 at the level of
significance 𝛼 = 0.05. Stated more completely, 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 exceeds 50 centimeters per second.
Alternatively, we can also use P-value approach for the decision.

P-value Approach (Using z-table)
9
𝑝 = 2 ∗ 𝑃 𝑍 ≥ 𝑧0 = 2 ∗ 𝑃 𝑍 ≥ 3.25 = 2 ∗ 1 − 𝑃 𝑍 < 3.25 = 2 ∗ 1 − Φ 3.25 = 2 ∗ 1 − 0.99918
𝑝 = 0.0016, since 𝑝 = 0.0016 < 0.05 = 𝛼, we reject the null hypothesis 𝐻0 : 𝜇 = 50 at the level of
significance 𝛼 = 0.05.
Example 2:
A melting point test of 𝑛 = 10 samples of a binder used in manufacturing a rocket propellant
resulted in 𝑥 = 154.2 𝑜 𝐹 Assume that melting point is normally distributed with 𝜎 = 1.5 𝑜 𝐹.
(a) 𝐻0 : 𝜇 = 155 𝑣𝑒𝑟𝑠𝑢𝑠 𝐻1 : 𝜇 ≠ 155 using 𝛼 = 0.01.
(b) What is the P-value for this test?
Example 3:
The life in hours of a battery is known to be approximately normally distributed, with standard
deviation 𝜎 = 1.25 hours. A random sample of 10 batteries has a mean life of 𝑥 = 40.5 hours.
(a) Is there evidence to support the claim that battery life exceeds 40 hours? Use 𝛼 = 0.05.
(b) What is the P-value for the test in part (a)?
Example 4:
Supercavitation is a propulsion technology for undersea vehicles that can greatly increase their
speed. It occurs above approximately 50 meters per second, when pressure drops sufficiently to
allow the water to dissociate into water vapor, forming a gas bubble behind the vehicle. When
the gas bubble completely encloses the vehicle, supercavitation is said to occur. Eight tests were
conducted on a scale model of an undersea vehicle in a towing basin with the average observed
speed 𝑥 = 102.2 meters per second. Assume that speed is normally distributed with known
standard deviation 𝜎 = 4 meters per second. Test the hypothesis 𝐻0 : 𝜇 = 100 𝑣𝑒𝑟𝑠𝑢𝑠 𝐻1 : 𝜇 >
100 using 𝛼 = 0.05.
Test of Hypothesis on mean 𝝁 on a Single Sample with unknown variance 𝝈𝟐

1. Test of Hypothesis on the mean of a Normal Distribution of unknown Variance - Two
tail test
Let us consider the given population X normally distributed with unknown mean 𝜇 but unknown
variance 𝜎 2
Alternate hypothesis: 𝐻1 : 𝜇 ≠ 𝜇0
10
0.05 𝑜𝑟 𝛼 = 0.01. (If you are using critical value approach for decision compute 𝑡𝛼/2,𝑛−1
𝑢𝑠𝑖𝑛𝑔 𝑡 − 𝑡𝑎𝑏𝑙𝑒).
1 𝑛 1 𝑛
average 𝑋 = 𝑛 𝑖=1 𝑋𝑖 and sample variance 𝑆 2 = 𝑛 −1 𝑖=1(𝑋𝑖 − 𝑋)2 . For this case we compute
𝑋 −𝜇
the test statistic 𝑡0 = 𝑆/√𝑛 . (If you want to use P-value approach, calculate the probability 𝑝 =
2 ∗ 𝑃 𝑡 ≥ 𝑡0 , t-distribution with n-1 degree of freedom 𝑢𝑠𝑖𝑛𝑔 𝑡 − 𝑡𝑎𝑏𝑙𝑒).
 Reject the null hypothesis 𝐻0 if either 𝑡0 > 𝑡𝛼 /2,𝑛−1 or 𝑡0 < 𝑡𝛼/2,𝑛−1 (𝑜𝑟 𝑡0 > 𝑡𝛼 /2,𝑛−1 )
 Fail to reject the null hypothesis 𝐻0 if−𝑡𝛼/2,𝑛−1 ≤ 𝑡0 < 𝑡𝛼 /2,𝑛−1

2. Test of Hypothesis on the mean of a Normal Distribution of unknown Variance - one

tail test (Right tail)
variance 𝜎 2
Step1: (State the Null and Alternative hypothesis)
Alternate hypothesis: 𝐻1 : 𝜇 > 𝜇0
11
0.05 𝑜𝑟 𝛼 = 0.01. (If you are using critical value approach for decision
compute 𝑡𝛼 ,𝑛−1 𝑢𝑠𝑖𝑛𝑔 𝑡 − 𝑡𝑎𝑏𝑙𝑒).
1 𝑛 1 𝑛
𝑋 −𝜇
𝑃 𝑡 ≥ 𝑡0 , t-distribution with n-1 degree of freedom).
 Reject the null hypothesis 𝐻0 if 𝑡0 > 𝑡𝛼,𝑛−1

 Fail to reject the null hypothesis 𝐻0 if 𝑡0 ≤ 𝑡𝛼,𝑛−1

3. Test of Hypothesis on the mean of a Normal Distribution of unknown Variance - one

tail test (Right tail)
variance 𝜎 2
Step1: (State the Null and Alternative hypothesis)
Alternate hypothesis: 𝐻1 : 𝜇 < 𝜇0
0.05 𝑜𝑟 𝛼 = 0.01. (If you are using critical value approach for decision
compute 𝑡𝛼 ,𝑛−1, 𝑢𝑠𝑖𝑛𝑔 𝑡 − 𝑡𝑎𝑏𝑙𝑒).
12
1 𝑛 1 𝑛
𝑋 −𝜇
𝑃 𝑡 ≥ 𝑡0 , t-distribution with n-1 degree of freedom 𝑢𝑠𝑖𝑛𝑔 𝑡 − 𝑡𝑎𝑏𝑙𝑒).
 Reject the null hypothesis 𝐻0 if 𝑡0 < −𝑡𝛼,𝑛−1

 Fail to reject the null hypothesis 𝐻0 if 𝑡0 ≥ −𝑡𝛼,𝑛−1

Example 1 (One tail t- test)

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 outgoing velocity of the ball 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 air cannon so that the incoming velocity and spin rate of the ball could be
precisely controlled. It is of interest to determine if there is evidence (with 𝛼 = 0.05) to support
a claim that the mean coefficient of restitution exceeds 0.82. 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
13
1 𝑛
Solution: The sample mean 𝑥 = 𝑛 𝑖=1 𝑥𝑖 = 0.83725 and sample standard deviation 𝑠 =
1 𝑛 2
𝑖=1 𝑥𝑖 − 𝑥 = 0.02456 from the sample size 𝑛 = 15. Since the objective of the
𝑛−1
experimenter is to demonstrate that the mean coefficient of restitution exceeds 0.82, a one-sided
alternative hypothesis is appropriate.
Null Hypothesis, 𝐻0 : 𝜇 = 0.82
Alternative Hypothesis, 𝐻1 : 𝜇 > 0.82
Level of significance 𝛼 = 0.05. Since sample variance is not known, we have to use one-sided
right tail t-test, we need 𝑡𝛼,𝑛−1 = 𝑡0.05,14 = 1.761
We have sample standard deviation 𝑠 = 0.02456, sample mean 𝑥 = 0.83725 , sample size 𝑛 =
15 and 𝜇0 = 0.82. The test statistics is
𝑥 − 𝜇0 0.83725 − 0.82
𝑡0 = = = 2.72
𝑠/√𝑛 2/√25

Since 𝑡0 = 2.72 > 1.761 = 𝑡𝛼,𝑛−1 , we reject the null hypothesis 𝐻0 : 𝜇 = 0.82 at the level of
significance 𝛼 = 0.05 that that the mean coefficient of restitution exceeds 0.82.

P-value Approach (Using t-table for 14 degrees of freedom)
We have, 𝑝 = 𝑃 𝑡 ≥ 𝑡0 = 𝑃 𝑡 ≥ 2.72 ≈ 0.01

𝑝 = 0.01, since 𝑝 = 0.01 < 0.05 = 𝛼, we reject the null hypothesis 𝐻0 : 𝜇 = 50 at the level of
significance 𝛼 = 0.05.
Example 2:
A 1992 article in the Journal of the American Medical Association (―A Critical Appraisal of 98.6
Degrees F, the Upper Limit of the Normal Body Temperature, and Other Legacies of Carl
Reinhold August Wundrlich‖) reported body temperature, gender, and heart rate for a number of
subjects. The body temperatures for 25 female subjects follow:
97.8, 97.2, 97.4, 97.6, 97.8,
97.9, 98.0, 98.0, 98.0, 98.1,
98.2, 98.3, 98.3, 98.4, 98.4,
98.4, 98.5, 98.6, 98.6, 98.7,
98.8, 98.8, 98.9, 98.9, 99.0.
Test the hypotheses 𝐻0 : 𝜇 = 98.6 𝑣𝑒𝑟𝑠𝑢𝑠 𝐻1 : 𝜇 ≠ 98, using 𝛼 = 0.05 Also, find the P-value
14
Example 3:
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 by Simpson, Alsen, and Eden,
―A Bayesian Analysis of a Multiplicative Treatment Effect in Weather Modification‖, 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.
Can you support a claim that mean rainfall from seeded clouds exceeds 25 acre-feet? Use 𝛼 =
0.05. What is P-value the statistics?
Test of Hypothesis on the Variance 𝝈𝟐 of a Normal Distribution

Let us consider a population X distributed normally with mean 𝜇 and variance 𝜎 2 (both
unknown), i.e. 𝑋~𝑁(𝜇, 𝜎 2 ). Let 𝑋1 , 𝑋2 , … , 𝑋𝑛 is random sample taken from the population and
1 𝑛
sample variance 𝑆 2 = 𝑛 −1 𝑖=1(𝑋𝑖 − 𝑋 )2 .
1. Two tail test
Step 1: (State the null and alternative hypotheses)

Null hypothesis: 𝐻0 : 𝜎 2 = 𝜎02
Alternate hypothesis: 𝐻1 : 𝜎 2 ≠ 𝜎02
where 𝜎02 is a specified constant.
0.05 𝑜𝑟 𝛼 = 0.01. For the critical value approach for decision compute 𝜒 2 𝛼/2,𝑛−1
and 𝜒 2 1−𝛼/2,𝑛−1 𝑢𝑠𝑖𝑛𝑔 𝜒 2 − 𝑡𝑎𝑏𝑙𝑒.

(𝑛−1)𝑆 2
For this case we compute the test statistic 𝜒0 2 = .
𝜎02
 Reject the null hypothesis 𝐻0 if either 𝜒0 2 > 𝜒 2 𝛼/2,𝑛−1 or 𝜒0 2 < 𝜒 2 1−𝛼/2,𝑛−1

 Fail to reject the null hypothesis 𝐻0 if 𝜒 2 1−𝛼/2,𝑛−1 ≤ 𝜒0 2 ≤ 𝜒 2 𝛼/2,𝑛−1
15
Where 𝑝 = 2 ∗ 𝑃(𝜒 2 > 𝜒 2 𝛼/2,𝑛−1 ) if 𝜒0 2 > 𝜒 2 𝛼 /2,𝑛−1 or 𝑝 = 2 ∗ 𝑃(𝜒 2 < 𝜒 2 1−𝛼/2,𝑛−1 ) if

𝜒0 2 < 𝜒 2 1−𝛼/2,𝑛−1, 𝑐𝑜𝑚𝑝𝑢𝑡𝑒 𝑢𝑠𝑖𝑛𝑔 𝜒 2 − 𝑡𝑎𝑏𝑙𝑒
2. One- tail test (Right tail test)

Null hypothesis: 𝐻0 : 𝜎 2 = 𝜎02 (𝜎 2 ≤ 𝜎02 )
Alternate hypothesis: 𝐻1 : 𝜎 2 > 𝜎02
0.05 𝑜𝑟 𝛼 = 0.01. For the critical value approach for decision compute 𝜒 2 𝛼,𝑛−1 𝑢𝑠𝑖𝑛𝑔 𝜒 2 −
𝑡𝑎𝑏𝑙𝑒.

(𝑛−1)𝑆 2
𝜎02
 Reject the null hypothesis 𝐻0 if either 𝜒0 2 > 𝜒 2 𝛼,𝑛−1

 Fail to reject the null hypothesis 𝐻0 if 𝜒0 2 ≤ 𝜒 2 𝛼 ,𝑛−1

Where 𝑝 = 𝑃(𝜒 2 > 𝜒 2 𝛼,𝑛−1 ) for 𝜒0 2 > 𝜒 2 𝛼,𝑛−1 , compute 𝑢𝑠𝑖𝑛𝑔 𝜒 2 − 𝑡𝑎𝑏𝑙𝑒
3. One- tail test (left-tail test)
Step 1: (State the null and alternative hypothesis)

Null hypothesis: 𝐻0 : 𝜎 2 = 𝜎02 (𝜎 2 ≥ 𝜎02 )
16
Alternate hypothesis: 𝐻1 : 𝜎 2 < 𝜎02
0.05 𝑜𝑟 𝛼 = 0.01. For the critical value approach for decision compute 𝜒 2 1−𝛼,𝑛−1 .

(𝑛−1)𝑆 2
𝜎02
 Reject the null hypothesis 𝐻0 if either 𝜒0 2 < 𝜒 2 𝛼−1,𝑛−1

 Fail to reject the null hypothesis 𝐻0 if 𝜒0 2 ≥ 𝜒 2 𝛼 −1,𝑛−1

Where 𝑝 = 𝑃(𝜒 2 < 𝜒 2 𝛼−1,𝑛−1 ) for 𝜒0 2 < 𝜒 2 𝛼−1,𝑛−1
Example 1:
An automatic filling machine is used to fill bottles with liquid detergent. A random sample of 20
bottles results in a sample variance of fill volume of 𝑠 2 = 0.0153 𝑓𝑙𝑢𝑖𝑑 𝑜𝑢𝑛𝑐𝑒𝑠 2 . If the
variance of fill volume exceeds 0.01 𝑓𝑙𝑢𝑖𝑑 𝑜𝑢𝑛𝑐𝑒𝑠 2 , an unacceptable proportion of bottles
will be under-filled or overfilled. Is there evidence in the sample data to suggest that the
manufacturer has a problem with under-filled or overfilled bottles? Use 𝛼 = 0.05, and assume
that fill volume has a normal distribution.
Solution: The parameter of interest is the population variance 𝜎 2

Null Hypothesis, 𝐻0 : 𝜎 2 = 0.01
Alternative Hypothesis, 𝐻1 : 𝜎 2 > 0.01
17
Level of significance 𝛼 = 0.05. We have to use one-sided right tail 𝜒 2 , we need 𝜒 2 𝛼,𝑛−1 =
𝜒 2 0.05,19 = 30.14
We have sample standard variance 𝑠 2 = 0.0153, sample size 𝑛 = 20 and 𝜎02 = 0.01. The test
statistics is
2
(𝑛 − 1)𝑠 2 19(0.0153)
𝜒0 = = = 29.07
𝜎02 0.01

Since 𝜒0 2 = 29.07 < 𝜒 2 0.05,19 = 30.14, we conclude that there is no strong evidence that the
variance of fill volume exceeds 0.01 𝑓𝑙𝑢𝑖𝑑 𝑜𝑢𝑛𝑐𝑒𝑠 2 . Fail to reject 𝐻0 .
P-value Approach (Using 𝜒 2 -table with 19 degrees of freedom)
We have, 𝑝 = 𝑃 𝜒 2 ≥ 𝜒0 2 = 𝑃 𝜒 2 ≥ 29.07 ≈ 0.0649

𝑝 = 0.0649, since 𝑝 = 0.0649 > 0.05 = 𝛼, we conclude that there is no strong evidence that
the variance of fill volume exceeds 0.01 𝑓𝑙𝑢𝑖𝑑 𝑜𝑢𝑛𝑐𝑒𝑠 2 .
Example 2:
The sugar content of the syrup in canned peaches is normally distributed A random sample of
n=10 cans yields a sample standard deviation of 𝑠 = 4.8 milligrams. Suppose that the variance
is thought to be 𝜎 2 = 18 milligrams 2 .
(a) Test the hypothesis 𝐻0 : 𝜎 2 = 18 𝑣𝑒𝑟𝑠𝑢𝑠 𝐻1 : 𝜎 2 ≠ 18 . Using 𝛼 = 0.05 .
Two Sample Test

1. Test of Hypothesis on the means 𝝁𝟏 𝒂𝒏𝒅 𝝁𝟐 of two normally distributed
populations with known variances
Let us consider two populations 𝑋1 𝑎𝑛𝑑 𝑋2 normally distributed such that 𝑋1 ~𝑁(𝜇1 , 𝜎12 ) and
𝑋2 ~𝑁(𝜇2 , 𝜎22 ) with known variances 𝜎12 𝑎𝑛𝑑 𝜎22 . Suppose that random sample 𝑋11 , 𝑋12 , … , 𝑋1𝑛 1
1 𝑛1
of size 𝑛1 is drawn from the population 𝑋1 with sample mean 𝑋1 = 𝑛 𝑖=1 𝑋1𝑖 and random
1
sample 𝑋21 , 𝑋22 , … , 𝑋2𝑛 2 of size 𝑛2 is drawn from the population 𝑋2 with sample mean 𝑋2 =
1 𝑛2
𝑛2 𝑖=1 𝑋2𝑖 .
Two-Sided Test
18
Null hypothesis: 𝐻0 : 𝜇1 = 𝜇2
Alternate hypothesis: 𝐻1 : 𝜇1 ≠ 𝜇2
0.05 𝑜𝑟 𝛼 = 0.01. (If you are using critical value approach for decision compute 𝑧𝛼/2 𝑢𝑠𝑖𝑛𝑔 𝑧 −
Table for 𝒛𝜶/𝟐
𝜶 𝒛𝜶/𝟐
0.1 𝒛𝟎.𝟎𝟓 = 𝟏. 𝟔𝟒
0.05 𝒛𝟎.𝟎𝟐𝟓 = 𝟏. 𝟗𝟔
0.01 𝒛𝟎.𝟎𝟎𝟓 = 𝟐. 𝟓𝟖

𝑋1 −𝑋2
For this case we compute the test statistic 𝑧0 = . (If you want to use P-value approach,
𝜎2 2
1 +𝜎 2
𝑛1 𝑛2
calculate the probability 𝑝 = 2 ∗ 𝑃 𝑍 ≥ 𝑧0 𝑢𝑠𝑖𝑛𝑔 𝑧 − 𝑡𝑎𝑏𝑙𝑒).
 Reject the null hypothesis 𝐻0 if either 𝑧0 > 𝑧𝛼/2 or 𝑧0 < −𝑧𝛼/2

 Fail to reject the null hypothesis 𝐻0 if −𝑧𝛼/2 ≤ 𝑧0 ≤ 𝑧𝛼/2

Note: One sided test (left and right tail tests) for the test of hypothesis of means of two
normally distributed populations with known variances can be done similarly.
Example 1
A product developer is interested in reducing the drying time of a primer paint. Two
formulations of the paint are tested; formulation 1 is the standard chemistry, and formulation 2
19
has a new drying ingredient that should reduce the drying time. From experience, it is known that
the standard deviation of drying time is 8 minutes, and this inherent variability should be
unaffected by the addition of the new ingredient. Ten specimens are painted with formulation 1,
and another 10 specimens are painted with formulation 2; the 20 specimens are painted in
random order. The two sample average drying times are 𝑥1 = 121 minutes and 𝑥2 =
112 minutes, respectively. What conclusions can the product developer draw about the
effectiveness of the new ingredient, using 𝛼 = 0.05?
Solution: The quantity of interest is the difference in mean dying times, 𝜇1 − 𝜇2 .
Alternate hypothesis: 𝐻1 : 𝜇1 > 𝜇2
Level of significance 𝛼 = 0.05. We need the critical value 𝑧𝛼 = 𝑧0.05 = 1.64
We are given 𝜎12 = 𝜎22 = (8)2 = 64 (Known), 𝑛1 = 𝑛2 = 10 and 𝑥1 = 121, 𝑥2 = 112. For this
𝑋1 −𝑋2 121−112
case we compute the test statistic 𝑧0 = = 64 64
= 2.52
𝜎2 2
1 +𝜎 2 +
10 10
𝑛1 𝑛2
Since 𝑧0 = 2.52 > 𝑧𝛼 = 1.64, we reject the null hypothesis 𝐻0 : 𝜇1 = 𝜇2 at 𝛼 = 0.05 the level
of significance and conclude that adding the new ingredient to the paint significantly reduces the
drying time.
P-value Approach (Using z-table)
𝑝 = 𝑃 𝑍 ≥ 𝑧0 = 𝑃 𝑍 ≥ 2.52 = 1 − 𝑃 𝑍 < 2.52 = 1 − Φ 2.52 = 0.0059

Since 𝑝 = 0.0016 < 0.05 = 𝛼, we reject the null hypothesis 𝐻0 : 𝜇1 = 𝜇2 at 𝛼 = 0.05 the level
of significance and Conclude that adding the new ingredient to the paint significantly reduces the
drying time.
Example :2
Two machines are used for filling plastic bottles with a net volume of 16.0 ounces. The fill
volume can be assumed normal, with standard deviation 𝜎1 = 0.020 and 𝜎2 = 0.025 ounces. A
member of the quality engineering staff suspects that both machines fill to the same mean net
20
volume, whether or not this volume is 16.0 ounces. A random sample of 10 bottles is taken from
the output of each machine.
Machine 1 Machine 2
16.03 16.01 16.02 16.03
16.04 15.96 15.97 16.04
16.05 15.98 15.96 16.02
16.05 16.02 16.01 16.01
16.02 15.99 15.99 16.00
(a) Do you think the engineer is correct? Use 𝛼 = 0.05.

Example 3:
Two types of plastic are suitable for use by an electronics component manufacturer. The
breaking strength of this plastic is important. It is known that 𝜎1 = 𝜎2 = 1.0 psi. From a random
sample of size 𝑛1 = 10 and 𝑛2 = 10, we obtain
𝑥1 = 162.5, 𝑥2 = 155.0 . The company will not adopt plastic 1 unless its mean breaking strength
exceeds that of plastic 2 by at least 10 psi. Based on the sample information, should it use plastic
1? Use 𝛼 = 0.05 reaching a decision.
2. Test of Hypothesis on the means 𝝁𝟏 𝒂𝒏𝒅 𝝁𝟐 of two normally distributed

populations with unknown variances
Let us consider two populations 𝑋1 𝑎𝑛𝑑 𝑋2 normally distributed such that 𝑋1 ~𝑁(𝜇1 , 𝜎12 ) and
𝑋2 ~𝑁(𝜇2 , 𝜎22 ) with unknown variances 𝜎12 𝑎𝑛𝑑 𝜎22 . Suppose that random sample
𝑋11 , 𝑋12 , … , 𝑋1𝑛 1 of size 𝑛1 is drawn from the population 𝑋1 with sample mean 𝑋1 =
1 𝑛1 1 𝑛1
𝑖=1 𝑋1𝑖 and sample variance 𝑆12 = 𝑛 𝑖=1(𝑋1𝑖 − 𝑋1 )2 and random sample
𝑛1 1 −1
𝑋21 , 𝑋22 , … , 𝑋2𝑛 2 of size 𝑛2 is drawn from the population 𝑋2 with sample mean 𝑋2 =
1 𝑛2 1 𝑛2
𝑖=1 𝑋2𝑖 and sample variance 𝑆22 = 𝑛 𝑖=1(𝑋1𝑖 − 𝑋2 )2 .
𝑛2 2 −1
There are two possible cases:

Case I: if 𝝈𝟐𝟏 = 𝝈𝟐𝟐 = 𝝈𝟐
Alternate hypothesis: 𝐻1 : 𝜇1 ≠ 𝜇2 (Two-Sided Test)
0.05 𝑜𝑟 𝛼 = 0.01. For critical value approach for decision, find 𝑡𝛼/2,𝑛 1 +𝑛 2 −2 𝑢𝑠𝑖𝑛𝑔 𝑡 − 𝑡𝑎𝑏𝑙𝑒.
21
𝑋1 −𝑋2 𝑛 1 −1 𝑆12 + 𝑛 2 −1 𝑆22

For this case we compute the test statistic 𝑡0 = , where 𝑆𝑃2 = is
𝑆𝑝
1
+
1 𝑛 1 +𝑛 2 −2
𝑛1 𝑛2
combined (pooled) sample variance
(If you want to use P-value approach, calculate the probability 𝑝 = 2 ∗ 𝑃 𝑡 ≥ 𝑡0 , 𝑡 −

𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛 𝑤𝑖𝑡𝑕 𝑛1 + 𝑛2 − 2 𝑑𝑒𝑔𝑟𝑒𝑒 𝑜𝑓 𝑓𝑟𝑒𝑒𝑑𝑜𝑚).
 Reject the null hypothesis 𝐻0 if either 𝑡0 > 𝑡𝛼 /2,𝑛 1 +𝑛 2 −2 or 𝑡0 < − 𝑡𝛼/2,𝑛 1 +𝑛 2 −2

 Fail to reject the null hypothesis 𝐻0 if − 𝑡𝛼/2,𝑛 1 +𝑛 2 −2 ≤ 𝑡0 ≤ 𝑡𝛼 /2,𝑛 1 +𝑛 2 −2

normally distributed populations with unknown variances can be done similarly.
Example 1:
Two catalysts are being analyzed to determine how they affect the mean yield of a chemical
process. Specifically, catalyst 1 is currently in use, but catalyst 2 is acceptable. Since catalyst 2 is
cheaper, it should be adopted, providing it does not change the process yield. A test is run in the
pilot plant and results in the data shown in Table below. Is there any difference between the
mean yields? Use 𝛼 = 0.05, and assume equal variances.
Table: Catalyst Yield Data
Observation Number Catalyst 1 Catalyst 2

1 91.50 89.19
2 94.18 90.95
3 92.18 90.46
4 95.39 93.21
5 91.79 97.19
6 89.07 97.04
7 94.72 91.07
8 89.21 92.7
Solution: The parameters of interest are and 𝜇1 and 𝜇2 , the mean process yield using catalysts
1 and 2, respectively, and we want to know if 𝜇1 = 𝜇2 .
22
Alternate hypothesis: 𝐻1 : 𝜇1 ≠ 𝜇2
Level of significance 𝛼 = 0.05. We need the critical value 𝑡𝛼/2,𝑛−1 = 𝑡0.025,7 = 2.145
Assumption is that 𝜎12 = 𝜎22 (Unknown). We are given 𝑛1 = 8, 𝑥1 = 92.255, 𝑠1 = 2.39, 𝑎𝑛𝑑 𝑛2 =
8, 𝑥2 = 92.733, 𝑠2 = 2.98.
We also need the combined (Pooled) sample variance

𝑛 1 −1 𝑠12 + 𝑛 2 −1 𝑠22 7×(2.39)2 +7×(2.98)2
𝑠𝑝2 = = = 7.30 , so 𝑠𝑝 = 2.70.
𝑛 1 +𝑛 2 −2 8+8−2
𝑥 1 −𝑥 2 92.255−92.733
For this case we compute the test statistic 𝑡0 = 1 1
= = −0.35
1 1
𝑠𝑝 + 2.70 +
𝑛1 𝑛2 8 8
Since 𝑡0 = −0.35 > −2.145 = −𝑡𝛼/2,𝑛−1 , the null hypothesis 𝐻0 cannot be rejected. That is,
𝛼 = 0.05 level of significance, we do not have strong evidence to conclude that catalyst 2 results
in a mean yield that differs from the mean yield when catalyst 1 is used.
P-value Approach (Using t-table at 7 degrees of freedom)
𝑝 = 2 ∗ 𝑃 𝑡 ≤ 𝑡0 = 𝑃 𝑡 ≤ −0.35 = 𝑃 𝑡 ≥ 0.35 , from the t-table we can see 𝑡0.40,14 =

0.258 𝑎𝑛𝑑 𝑡0.25,14 = 0.692. Therefore, 0.258 < 0.35 < 0.692 ⇒ 0.50 < 𝑝 < 0.80 which
exceed 𝛼 = 0.05, the null hypothesis cannot be rejected. Thus, we do not have strong evidence
to conclude that catalyst 2 results in a mean yield that differs from the mean yield when catalyst
1 is used.
Example:2
The diameter of steel rods manufactured on two different extrusion machines is being
investigated. Two random samples of sizes 𝑛1 = 15 and 𝑛2 = 17 are selected, and the sample
means and sample variances are 𝑥1 = 8.73, 𝑠12 = 0.35, 𝑎𝑛𝑑 𝑥2 = 8, 𝑥2 = 8.68, 𝑠22 = 0.40,
respectively. Assume that 𝜎12 = 𝜎22 and that the data are drawn from a normal distribution.
(a) Is there evidence to support the claim that the two machines produce rods with different
mean diameters? Use 𝛼 = 0.05 in arriving at this conclusion.
(b) Find the P-value for the t-statistic you calculated in part (a).
(c) Construct a 95% confidence interval for the difference in mean rod diameter. Interpret this
interval.
23
Example 3:
Two catalysts may be used in a batch chemical process. Twelve batches were prepared using
catalyst 1, resulting in an average yield of 86 and a sample standard deviation of 3. Fifteen
batches were prepared using catalyst 2, and they resulted in an average yield of 89 with a
standard deviation of 2. Assume that yield measurements are approximately normally distributed
with the same standard deviation. (a) Is there evidence to support a claim that catalyst 2 produces
a higher mean yield than catalyst 1? Use = 0.01 . (b) Find a 95% confidence interval on the
difference in mean yields.
Case II: if 𝝈𝟐𝟏 ≠ 𝝈𝟐𝟐

0.05 𝑜𝑟 𝛼 = 0.01. For critical value approach for decision, find 𝑡𝛼/2,𝜈 , where degrees of freedom
𝜈 is defined below using 𝑡 − 𝑡𝑎𝑏𝑙𝑒.

𝑥 1 −𝑥 2
For this case we compute the test statistic 𝑡0 = , which is t-distributed with degrees of
𝑠2
1 𝑠2
2
+
𝑛1 𝑛2
2
𝑠 21 𝑠 22
+
𝑛1 𝑛2
freedom 𝜈 = 2 2 −2
𝑠 21 𝑠 22
𝑛1 𝑛2
+
𝑛 1 −1 𝑛 2 −1
(If you want to use P-value approach, calculate the probability 𝑝 = 2 ∗ 𝑃 𝑡 ≥ 𝑡0 , t-distribution
with 𝜈 degrees of freedom. Compute using t-table)
 Reject the null hypothesis 𝐻0 if either 𝑡0 > 𝑡𝛼/2,𝜈 or 𝑡0 < − 𝑡𝛼/2,𝜈

 Fail to reject the null hypothesis 𝐻0 if − 𝑡𝛼/2,𝜈 ≤ 𝑡0 ≤ 𝑡𝛼/2,𝜈
24
 Fail to reject the null hypothesis 𝐻0 if 𝑝 ≥ 𝛼
normally distributed populations with unknown variances can be done similarly.
Example 1:
Arsenic concentration in public drinking water supplies is a potential health risk. A research
article reported drinking water arsenic concentrations in parts per billion (ppb) for 10 village
communities in Eastern Tarai region and 10 village communities in Western Tarai region. The
data follow:
Eastern Region Arsenic Western Region Arsenic
Community 1 3 Community 1 48
Determine is there any difference in mean arsenic concentrations between Eastern Regional
communities and communities in the Western region? Use 𝛼 = 0.05. We assume each
population is normally distributed but that their variances are not equal. What is the P-value for
this test?
Solution: The parameters of interest are the mean arsenic concentrations for the two geographic
regions, say, 𝜇1 (Eastern region community) and 𝜇2 (Western region community), and we are
interested in determining whether 𝜇1 = 𝜇2 .
From the given table, we calculated that 𝑥1 = 12.5, 𝑠12 = 58.22, 𝑛1 = 10, for data in Eastern
region community and 𝑥2 = 27.5, 𝑠22 = 234.09, 𝑛2 = 10 for data in Western region community
25
Level of significance 𝛼 = 0.05. Since the variances 𝜎12 𝑎𝑛𝑑 𝜎22 are unknown and 𝜎12 =
𝑠2 2 2
1 + 𝑠2
𝑛1 𝑛2
𝜎22 . We have to apply t-test with degrees of freedom given by 𝜈 = 2 2 −2 =
𝑠2
1 𝑠2
2
𝑛1 𝑛2
+
𝑛 1 −1 𝑛 2 −1
58.22 234.09 2
+
10 10
58.22 2 234.09 2
− 2 = 13.2 ≈ 13. We need 𝑡𝛼/2,𝜈 = 𝑡0.025,13 = 2.16 for critical value test.
10
+ 10
9 9

𝑥 1 −𝑥 2 12.5−27.5
For this case we compute the test statistic 𝑡0 = = = −2.77
58.22 234.09
𝑠2
1 𝑠2
2 +
+ 10 10
𝑛1 𝑛2
Since 𝑡0 = −2.77 < −2.16 = − 𝑡𝛼/2,𝜈 , we reject the null hypothesis 𝐻0 : 𝜇1 = 𝜇2 . Therefore,
there is evidence to conclude that mean arsenic concentration in the drinking water in rural
Eastern region is different from the mean arsenic concentration in Western region drinking
water.
The P-value for this test is 𝑝 = 2 ∗ 𝑃 𝑡 ≤ 𝑡0 = 2 ∗ 𝑃 𝑡 ≤ −2.77 = 2 ∗ 𝑃 𝑡 ≥ 2.77 ≈ 0.016
Example 2:
Two companies manufacture a rubber material intended for use in an automotive application.
The part will be subjected to abrasive wear in the field application, so we decide to compare the
material produced by each company in a test. Twenty-five samples of material from each
company are tested in an abrasion test, and the amount of wear after 1000 cycles is observed. For
company 1, the sample mean and standard deviation of wear are 𝑥1 =20
milligram/1000 𝑐𝑦𝑐𝑙𝑒𝑠 , 𝑠1 = 2 milligram/1000 𝑐𝑦𝑐𝑙𝑒𝑠, while for company 2 we obtain 𝑥1 =15
milligram/1000 𝑐𝑦𝑐𝑙𝑒𝑠 , 𝑠2 = 8 milligram/1000 𝑐𝑦𝑐𝑙𝑒𝑠.
(a) Do the data support the claim that the two companies produce material with different mean
wear? 𝛼 = 0.05, and assume each population is normally distributed but that their variances
are not equal.
Example 3:
The thickness of a plastic film (in mils) on a substrate material is thought to be influenced by the
temperature at which the coating is applied. A completely randomized experiment is carried out.
Eleven substrates are coated at 125 °𝐶, resulting in a sample mean coating thickness of and a
sample standard deviation of 𝑠1 = 10.2 . Another 13 substrates are coated at150 °𝐶, for which
26
and 𝑠2 = 20.1 are observed. It was originally suspected that raising the process temperature
would reduce mean coating thickness. Do the data support this claim? Use 𝛼 = 0.01 and assume
that the two population standard deviations are not equal. Calculate an approximate P-value for
this test.
Remark: For example we want to test the null hypothesis 𝐻0 : 𝜇1 − 𝜇2 = 𝛿, for some constant 𝛿,
against the alternative hypothesis 𝐻1 : 𝜇1 − 𝜇2 ≠ 𝛿, in above all test statistics the numerator
𝑥 1 −𝑥 2 −𝛿 𝑥 1 −𝑥 2 − 𝛿
expression is subtracted with 𝛿, that mean, 𝑧0 = for known variances, 𝑡0 = 1 1
for
𝜎2 2
1 +𝜎 2 𝑠𝑝 +
𝑛1 𝑛2
𝑛1 𝑛2
𝑥 1 −𝑥 2 −𝛿
unknown variances but assumed equal and 𝑡0 = for unknown variances and assumed
𝑠2 2
1 + 𝑠2
𝑛1 𝑛2
unequal. Other steps follow similar as explained above.
The Paired t-test
When two samples of the two populations are collected in a pair under
homogeneous conditions but conditions may change from one pair to another.
Let (𝑋11 , 𝑋21 ),(𝑋12 , 𝑋22 ), … ..,(𝑋1𝑛 , 𝑋2𝑛 ) be set of n- paired observations, where
𝑋1𝑖 ~ 𝑁 𝜇1 , 𝜎12 and 𝑋2𝑖 ~ 𝑁 𝜇2 , 𝜎22 , 𝑖 = 1,2, … , 𝑛. We want to test the hypothesis
about the equality of the means 𝜇1 and 𝜇2.
Notations: 𝜇𝐷 = 𝜇1 − 𝜇2 , 𝐷𝑖 = 𝑋1𝑖 − 𝑋2𝑖 , 𝑖 = 1, … , 𝑛.
Step1: (State the Null and Alternative hypotheses )
Null hypothesis: 𝐻0 : 𝜇𝐷 = 0
Alternate hypothesis: 𝐻1 : 𝜇𝐷 ≠ 0
0.05 𝑜𝑟 𝛼 = 0.01. (If you are using critical value approach for decision compute 𝑡𝛼/2,𝑛−1 using t-
table.)

1 𝑛
We compute the sample average of differences 𝐷 = 𝑛 𝑖=1 𝐷𝑖 and sample variance of the
1 𝑛 𝐷
differences 𝑆𝐷2 = 𝑛−1 𝑖=1 𝐷𝑖 − 𝐷 2
. For this case we compute the test statistic 𝑡0 = 𝑠 . (If
𝐷 /√𝑛
you want to use P-value approach, calculate the probability 𝑝 = 2 ∗ 𝑃(𝑡 ≥ 𝑡0 )) with t-
distribution with n-1 degrees of freedom. Compute using t-table.
27
 Reject the null hypothesis 𝐻0 if either 𝑡0 > 𝑡𝛼/2,𝑛−1 or 𝑡0 < −𝑡𝛼/2,𝑛−1

 Fail to reject the null hypothesis 𝐻0 if −𝑡𝛼/2,𝑛−1 ≤ 𝑡0 ≤ 𝑡𝛼/2,𝑛−1

 Fail to reject the null hypothesis 𝐻0 if 𝑝 ≥ 𝛼
Note: One sided hypothesis test (left and right tail tests) can be done similarly.
Example 1:
Ten individuals have participated in a diet-modification program to stimulate weight loss. Their
weight both before and after participation in the program is shown in the table below. Is there
evidence to support the claim that this particular diet-modification program is effective in
producing a mean weight reduction? Use 𝛼 = 0.05.
Subject Before After
1 195 187
2 213 195
3 247 221
4 201 190
5 187 175
6 210 197
7 215 199
8 246 221
9 294 278
10 310 285
Solution: The parameter of interest is the mean weight loss after participation in a diet-
modification program. From the given table, we tabulate the difference in the weight of the
individual:
𝐷𝑖 8 18 26 11 12 13 16 25 16 25
Sample mean and sample standard deviation of the difference
𝐷 = 17 𝑎𝑛𝑑 𝑠 = 6.4118, 𝑛 = 10. Notation: 𝜇𝐷 = 𝜇1 − 𝜇2

Null hypothesis: 𝐻0 : 𝜇𝐷 = 0
Alternate hypothesis: 𝐻1 : 𝜇𝐷 ≠ 0 (Two-Sided Test)
Level of significance 𝛼 = 0.05. We need 𝑡𝛼 /2,𝑛−1 = 𝑡0.025,9 = 2.262 for critical value test.
28
𝐷 17
For this case we compute the test statistic 𝑡0 = 𝑆 = 6.4118/√9 = 7.9541
𝐷 /√𝑛
Since 𝑡0 = 7.9541 > 2.262 = 𝑡𝛼/2,𝑛−1 , we reject the null hypothesis 𝐻0 : 𝜇𝐷 = 0. Therefore,
there is evidence to support the claim that this particular diet-modification program is effective in
producing a mean weight reduction.
The P-value for this test is 𝑝 = 2 ∗ 𝑃 𝑡 ≥ 𝑡0 = 2 ∗ 𝑃 𝑡 ≥ 7.9541 < 2 ∗ 0.0005 = 0.001

Example 2:
Two different analytical tests can be used to determine the impurity level in steel alloys. Eight
specimens are tested using both procedures, and the results are shown in the following
tabulation. Is there sufficient evidence to conclude that both tests give the same mean impurity
level, using 𝛼 = 0.01?
Specimen Test 1 Test 2
1 1.2 1.4
2 1.3 1.7
3 1.5 1.5
4 1.4 1.3
5 1.7 2.0
6 1.8 2.1
7 1.4 1.7
8 1.3 1.6
Test for the equality of two variances

Let us consider two populations 𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛1~𝑁(𝜇1 , 𝜎12 ) and𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛2~𝑁(𝜇2 , 𝜎22 ), where
𝜇1 , 𝜎12 , 𝜇2 , 𝜎22 are unknown. Given the sample 𝑋11 , 𝑋12, … , 𝑋1𝑛 1 of size 𝑛1 of 𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛1 with
1 𝑛1
sample variance 𝑠12 = 𝑛 𝑖=1(𝑥1𝑖 − 𝑋1 )2 and sample 𝑋21 , 𝑋22 , … , 𝑋2𝑛 2 of size 𝑛2 of the
1 −1
1 𝑛2
𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛2 with sample variance 𝑠22 = 𝑛 𝑖=1(𝑋2𝑖 − 𝑋2 )2 . We wish to test the hypothesis
2 −1
about the equality of the two variances 𝜎12 and 𝜎22

Null hypothesis: 𝐻0 : 𝜎12 = 𝜎22
Alternate hypothesis: 𝐻1 : 𝜎12 ≠ 𝜎22 (Two tail test)
29
0.05 𝑜𝑟 𝛼 = 0.01. For the critical value approach for decision, find 𝐹𝛼/2,𝑛 1 −1,𝑛 2 −1
and𝐹1−𝛼/2,𝑛 1 −1,𝑛 2 −1 using F-table.

𝑠1 2
For this case we compute the test statistic 𝐹0 = , which is distributed as F, with 𝑛1 −
𝑠2 2
1 and 𝑛2 − 1 degrees of freedom.
 Reject the null hypothesis 𝐻0 if either 𝐹0 > 𝐹𝛼 /2,𝑛 1 −1,𝑛 2 −1 or 𝐹0 < 𝐹1−𝛼/2,𝑛 1 −1,𝑛 2 −1
 Fail to reject the null hypothesis 𝐻0 if 𝐹1−𝛼/2,𝑛 1 −1,𝑛 2 −1 ≤ 𝐹0 ≤ 𝐹𝛼 /2,𝑛 1 −1,𝑛 2 −1

Where 𝑝 = 2 ∗ 𝑃(𝐹 > 𝐹𝛼/2,𝑛 1 −1,𝑛 2 −1 ) if 𝐹0 > 𝐹𝛼 /2,𝑛 1 −1,𝑛 2 −1 OR 𝑝 = 2 ∗ 𝑃(𝐹 <
𝐹1−𝛼/2,𝑛1−1,𝑛2−1) if 𝐹0<𝐹𝛼/2,𝑛1−1,𝑛2−1 . Compute using F- table.
Note: One sided hypothesis test (left and right tail tests) can be done similarly.
Example 1:
Oxide layers on semiconductor wafers are etched in a mixture of gases to achieve the proper
thickness. The variability in the thickness of these oxide layers is a critical characteristic of the
wafer, and low variability is desirable for subsequent processing steps. Two different mixtures of
gases are being studied to determine whether one is superior in reducing the variability of the
oxide thickness. Twenty wafers are etched in each gas. The sample standard deviations of oxide
thickness are 𝑠1 = 1.96 angstroms and 𝑠2 = 2.13 angstroms, respectively. Is there any evidence
to indicate that either gas is preferable? Use 𝛼 = 0.05
Solution: The parameters of interest are the variances of oxide thickness 𝑠12 and 𝑠22 . We will
assume that oxide thickness is a normal random variable for both gas mixtures.
Null hypothesis: 𝐻0 : 𝜎12 = 𝜎22
Alternate hypothesis: 𝐻1 : 𝜎12 ≠ 𝜎22 (Two tail test)
Level of significance 𝛼 = 0.05. For the critical value approach for decision, we need
𝐹𝛼/2,𝑛 1−1,𝑛 2−1 = 𝐹0.025,19,19 = 2.53 and 𝐹1−𝛼/2,𝑛 1−1,𝑛 2−1 = 𝐹0.975,19,19 = 0.40 (using F-table).
30
𝑠1 2 1.96 2
For this case we compute the test statistic 𝐹0 = 2
= = 0.85
𝑠2 2.13 2
Since 𝐹1−𝛼/2,𝑛 1 −1,𝑛 2−1 = 0.40 < 𝐹0 = 0.85 < 2.53𝐹𝛼/2,𝑛 1−1,𝑛 2 −1 , we cannot reject the null
hypothesis 𝐻0 : 𝜎12 = 𝜎22 at the 𝛼 = 0.05 level of significance. Therefore, there is no strong
evidence to indicate that either gas results in a smaller variance of oxide thickness.
Example 2:
Two chemical companies can supply a raw material. The concentration of a particular element in
this material is important. The mean concentration for both suppliers is the same, but we suspect
that the variability in concentration may differ between the two companies. The standard
deviation of concentration in a random sample of 𝑛1 = 10 batches produced by company 1 is
𝑠1 = 4.7 grams per liter, while for company 2, a random sample of 𝑛2 = 16 batches yields
𝑠2 = 5.8 grams per liter. Is there sufficient evidence to conclude that the two population
variances differ? Use 𝛼 = 0.05.
31

Test of Hypotheses: Hypothesis

Uploaded by

Copyright:

Available Formats

You might also like

Test of Hypotheses: Hypothesis

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Test of Hypotheses: Hypothesis

Uploaded by

Copyright:

Available Formats

Kathmandu University

Course: MATH 208, ENVE II/II

Prepared by: Dr. Samir Shrestha

The general idea of hypothesis testing

Example: Is normal body temperature really 98.6 degrees F?

Two Possibilities in Decision:

By Analyzing the evidence (observed data)

In this research question: (Constructing the Hypothesis)

𝑁𝑢𝑙𝑙 𝐻𝑦𝑝𝑜𝑡𝑕𝑒𝑠𝑖𝑠 𝐻0 : 𝜇 = 98.6

𝐴𝑙𝑡𝑒𝑟𝑛𝑎𝑡𝑒 𝐻𝑦𝑝𝑜𝑡𝑕𝑒𝑠𝑖𝑠 𝐻1 : 𝜇 ≠ 98.6

Example: Criminal Trial

𝑯𝟎 : Defendant is not guilty (innocent)

In statistics, the data are the evidence.

The jury then makes a decision based on the available evidence:

Jury Decision Not Guilty Guilty

Decision Null Hypothesis Alternative Hypothesis

Type I error: The null hypothesis is rejected when it is true.

Type II error: The null hypothesis is not rejected when it is false.

𝑃 𝑇𝑦𝑝𝑒 𝐼 𝑒𝑟𝑟𝑜𝑟 = 𝑃(𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑤𝑕𝑒𝑛 𝐻0 𝑖𝑠 𝑡𝑟𝑢𝑒 ) = 𝛼 (Called Level of significance)

𝑃 𝑇𝑦𝑝𝑒 𝐼𝐼 𝑒𝑟𝑟𝑜𝑟 = 𝑃(𝐹𝑎𝑖𝑙𝑒𝑑 𝑡𝑜 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 𝑤𝑕𝑒𝑛 𝐻0 𝑖𝑠 𝑓𝑎𝑙𝑠𝑒 ) = 𝛽

Making the decision (Two Approaches)

 Critical value approach (Used in many of the older textbooks). or

There are basically four steps to do Hypothesis testing:

Step 4: Make decision (Either by Critical Value Approach or P-Value Approach)

Two-Sided and One-Sided Hypotheses Examples

Null hypothesis: 𝐻0 : 𝜇 = 100

Alternate hypothesis: 𝐻1 : 𝜇 ≠ 100

Alternate hypothesis: 𝐻1 : 𝜇 ≠ 0.50

One-sided Hypothesis (right tail test)

Null hypothesis: 𝐻0 : 𝜇 = 100 (or 𝜇 ≤ 100)

Alternate hypothesis: 𝐻1 : 𝜇 > 100

Null hypothesis: 𝐻0 : 𝜇 = 10 (or 𝜇 ≤ 10)

Alternate hypothesis: 𝐻1 : 𝜇 > 10

One-sided Hypothesis (left tail test)

Null hypothesis: 𝐻0 : 𝜇 = 34 (or 𝜇 ≥ 34)

Alternate hypothesis: 𝐻1 : 𝜇 < 34

Single Sample Test

Step1: (State the Null and Alternative hypotheses)

where 𝜇0 is a specified constant.

Step2: (Provide the Level of significance 𝜶 )

Table for 𝒛𝜶/𝟐

Step3: (Test Statistics):

Step4: (Decision Making)

Critical Value Approach to make decision

(Equivalent) P-Value Approach to make decision

 Reject the null hypothesis 𝐻0 if 𝑝 < 𝛼

We are interested to test the following Hypothesis:

Alternate hypothesis: 𝐻1 : 𝜇 > 𝜇0

where 𝜇0 is a specified constant.

Step2: (Provide the Level of significance 𝜶 )

Step3: (Test Statistics):

Step4: (Decision Making)

Critical Value Approach to make decision

 Reject the null hypothesis 𝐻0 if either 𝑧0 > 𝑧𝛼

(Equivalent) P-Value Approach to make decision

 Reject the null hypothesis 𝐻0 if 𝑝 < 𝛼

We are interested to test the following Hypothesis: