SH1660

Test of Hypothesis

Statistical Hypothesis – is an assertion or conjecture concerning one or more populations.

Types of Statistical Hypotheses

· Null Hypothesis – denoted by ! , is a statement that shows that there is no significant
difference between population parameter and the value that is being claimed. It is the
hypothesis to be tested.
· Alternative Hypothesis – denoted by " or # , is a statement that shows that there is a
significant difference between the population parameter and the value that is being claimed.
This is a statement that will be true once the null hypothesis is rejected.

Null and Alternative Hypothesis Examples

1. The manager of a TV Station claims that the mean number of people in Manila watching their
new sitcom is 500,000 each day. A student researcher who wants statistical evidence on this
claim conducted her own survey. Her survey resulted to a mean of 499,995. Using a 0.05
significance level, can it be concluded that the mean number of people watching the new sitcom
is less than 500,000? Identify the null and the alternative hypotheses.
2. Last year, the mean number of bags produced by GB Company was 2,500 each day. This year,
the manager claims that there is an increase in the number of bags produced. A researcher who
wants to find out whether this is true counted the number of bags produced each day for a
period of one (1) month. His computation resulted to a mean of 2,515 bags. At 0.05 significance
level, is there enough evidence to conclude that the mean number of bags produced by GB
Company is greater than 2,500? Identify the null and the alternative hypothesis.
3. The Head of the Mathematics Department of a certain high school claims that the mean height
of Grade 7 students is 163 cm. The mean height of 45 randomly selected Grade 7 students is
161 cm. Using 0.01 significance level, can it be concluded that the mean height of Grade 7
students is different from 163 cm as claimed by the Head of the Mathematics Department?
Identify the null and the alternative hypotheses.

Types of Hypothesis Test

· Left-tailed - if # states that the parameter is less than the value claimed in ! .
· Right-tailed - if # states that the parameter is greater than the value claimed in !
· Two-tailed - if # states that the parameter is different from (or not equal to) the value claimed
in ! .

Test Statistic Sample Situation

Shennen owns a cat named Token. She brings Token to a local veterinarian for a regular check-up. Let
$ be a random variable that represents Token’s resting heart rate (in beats per minute). Based on past
records, $ has a normal distribution with % = 12. The vet checked the manual for this kind of cat and
found that for cats of this breed, & = 140 beats per minute. Over the past 5 weeks, Token’s heart rate
(beats/min) measured as follows: 147 126 138 119 108.
The sample mean $ = 127.6. The vet is concerned that Token’s heart rate may be slowing. Does the
data indicate that this is the case?

09 Handout 1 *Property of STI

Page 1 of 8
 SH1660

Answer: Since null hypothesis basically means “nothing has changed”, we can assume that
Token’s heart rate should be near average. Hence, ! : & = 140
Since the vet is concerned that Token’s heart rate is slowing, this point of view is represented by
the alternative hypothesis # : & < 140

Getting the probability of obtaining a sample mean of 127.6 or less from a population with true
mean µ = 140. If this probability is small, we can conclude that ! : & = 140 is not valid. Rather,
# : & < 140 and Token’s heart rate is slowing.
()* #./.3)#5!
Test statistic = ' = + = 67 ≈ −2.31
√- √8

Interpretation: If ! : µ = 140 is in fact true, the probability of getting a sample mean of $ ≤ 127.6
is only about 1.044% Because this probability is small, we reject ! : µ = 140 and conclude that
# : µ < 140. Token’s average heart rate seems to be slowing.

The E-value of a Hypothesis Test

The F-value or calculated probability is the estimated probability of rejecting the null hypothesis ( ! )
of a study question when that hypothesis is true. The smaller the F-value computed from sample data,
the stronger the evidence against ! .

E-value and the types of Hypothesis Test

Let G( represent the standardized sample test statistic for testing a mean & using the standard
()H
normal distribution. That is, G( = I .
K J
√
I. Left-tailed Test F-value = F(' < '( )
!: & = L #: & < L This is the probability of getting a test statistic
as low as or lower than '( .

II. Right-tailed Test F-value = F(' > '( )

!: & = L #: & > L This is the probability of getting a test statistic
as high as or higher than '( .

09 Handout 1 *Property of STI

Page 2 of 8
 SH1660

Q)RSTUV
III. Two-Tailed Test = F(' > |'( |); therefore,
!: & = L #: & ≠ L
.

F-value = 2F(' > |'( |)

This is the probability of getting a test statistic

either lower than −|'( | or higher than |'( |.

APPLICATION:
The Department of Environment and National Resources (DENR) has been studying Pringle Creek
regarding ammonia nitrogen concentration. The concentration has been 2.2 mg/l for couple of years.
However, the residents of the village where the creek is located are raising concern that the
concentration may have changed. Any change in the ammonia nitrogen concentration can affect the
living things such as plants and animals in and around the creek. Let $ be a random variable
representing ammonia nitrogen concentration. Based on the recent studies of Pringle Creek, $ has a
normal distribution with % = 0.30. A random sample of eight (8) water test from the creek is as follows:
2.1 2.0 2.5 2.2 2.7 2.9 2.1 2.4.

Construct a statistical test to examine the claim that the concentration of ammonia nitrogen has
changed from 2.2 mg/l. Use level of significance W = 0.01

Basic Components of a Statistical Test

a. Null Hypothesis ! , Alternative Hypothesis # , and preset level of significance α
If the evidence against ! is strong enough, we reject ! and adopt # . The level of significance
α is the probability of rejecting ! when it is, in fact, true.
b. Test Statistic and Sampling Distribution
These are the mathematical tools used to measure compatibility of sample data and the null
hypothesis.
c. F-value
This is the probability of obtaining a test statistic from the sampling distribution that is extreme as,
or more extreme than, the sample test statistic computed from the data under the assumption that
! is true.

d. Test Conclusion
If F-value ≤ α, we reject ! and say that the data are significant at level α. If F-value > α, we do
not reject ! .
e. Interpretation of the test results
Give a simple explanation of your conclusions in the context of the application.

Steps in Hypothesis Testing Example 1

Jack ‘n Jill, a company that produces snack foods, uses a machine to package 453-oz bags of potato
chips. Assuming that the net weights are normally distributed with mean 453oz with the standard

09 Handout 1 *Property of STI

Page 3 of 8
 SH1660

deviation of all such weights is 7.8oz, they want to test if the machine is working properly. A random
sample of 25 bags is checked. The weights are as follows:
464 457 438 453 448 449 440 450
469 432 454 464 449 446 445 455
446 458 456 434 451 447 448 446
452
Does the sample provide sufficient evidence to conclude that the packaging machine is not working
properly at a 5% level of significance?

Testing mean X when Y is unknown Example

The drug 6-mP is used to treat leukemia. The following data represent the remission times (in weeks)
for a random sample of 20 patients using 6-mP. (Brase, 2012)
9 8 31 23 21 5 17 33
32 26 19 20 7 17 36 7
12 9 6 10
The sample mean is $ ≈ 17.4 weeks with sample standard deviation Z ≈ 10.08. Let $ be a random
variable representing the remission time (in weeks) for all the patients using 6-mP. Assume that $
distribution is mound-shaped and symmetric.
A previously used drug treatment dah a mean remission time of µ = 12.5 weeks. Do the data indicate
that the mean remission time using the drug 6-mP is different (either way) from 12.5 week at a 1%
level of significance?

Process in Getting E-value

Given:
\ = 2.108
] = 21
^. _. = 21 − 1 = 20
Type of test = two-tailed test
Find the F-value.
Solution:

Using the table for \-distribution, we find an interval containing the F-value. Since this is a two-tailed
test, we use entries from the row headed by two-tail area. Look up the \-value in the row headed by
d. _ = 20. The sample statistic \ = 2.108 falls between 2.086 and 2.528. The F-value for the sample
t falls between the corresponding two-tail areas 0.050 and 0.020.

09 Handout 1 *Property of STI

Page 4 of 8
 SH1660

See the figure below for easier understanding.

Excerpt from Student’s t Distribution

... ...
one-tail area
0.050 0.020
two-tail area

^. _ = 20
2.086 2.528

Sample t = 2.108
E – Value:
0.020 < E-value < 0.050

Testing X Using Critical Regions

Figure 1. Critical Regions for `a : X = b

(a) # : & < L (b) # : & > L (c) # : & ≠ L

Left-tailed Right-tailed Two-tailed

W
W

Figure 2: Critical Values

Level of c = a. ad c = a. ae
Significance

09 Handout 1 *Property of STI

Page 5 of 8
 SH1660

For a left-tailed
test
`e : X < b
Critical Value fa
Critical Region:
All z < fa

For a right-tailed
test
`e : X > b
Critical Value fa
Critical Region:
All z > fa

For a two-tailed
test
`e : X ≠ b
Critical Value fa
Critical Region:
All z < −fa
together with all
z > +fa

Situation:
Let $ be a random variable representing the number of sunspots observed in a four-week period. A
random sample of 40 such periods from Spanish colonial times gave the number of sunspots per period.

The sample mean is $ ≈ 47.0 with % = 35. It is thought that for thousands of years, the mean number
of sunspots per four-week period was about & = 41. Do the data indicate that the mean sunspot activity
during the Spanish colonial period was higher than 41? Use W = 0.05. (Brase, 2012)

Population Proportion Formula

î − i î − i
'= mn ' =
il
k ki(1 − i)
] ]

09 Handout 1 *Property of STI

Page 6 of 8
 SH1660

where:
î = sample proportion
i = population proportion
] =sample size
l =1−i

The following are the steps when testing hypothesis concerning a proportion:
Step 1 – State the null and the alternative hypotheses.
Step 2 – Choose a level of significance W.
Step 3 – Compute the test statistic.
Step 4 – Determine the critical value.
Step 5 – Make a decision.

Null Hypothesis:
The null hypothesis is usually written in the form:
! : = i!
where i! = specific numerical value for the population proportion i.

Alternative Hypothesis:
The alternative hypothesis can be any of the following:
" : i ≠ i! (two-tailed test)
" : i > i! (one-tailed test)
" : i < i! (one-tailed test)

Example:
It is believed that in the coming election, 65% of the voters in the province of Pampanga will vote for
the administration candidate for governor. Suppose 713 out of the 1,150 randomly selected voters
indicate that they would vote for the administration candidate. At 0.10 level of significance, find out
whether the percentage of voters for the administration candidate is different from 65%.

Solution:
Step 1: State the null and the alternative hypotheses.
! : i = .65
" : i ≠ .65
Step 2: Choose a level of significance: W = 0.10

Step 3: Compute the test statistic.

$
î =
]
713
=
1,150
= 0.62

î − i
'=
ki(1 − i)
]

09 Handout 1 *Property of STI

Page 7 of 8
 SH1660

0.62 − 0.65
'=
0.65(1 − 0.65)
k
1,150
' = −2.133
Step 4: Determine the critical value.
The alternative hypothesis is non-directional. Hence, the two-tailed test shall be used. Divide W by 2,
and then subtract the quotient from 0.5.
W 0.10
= = 0.05
2 2
0.5 − 0.05 = 0.45
Using the Areas Under the Normal Curve Table, 'p = 1.645. At 10% level of significance the critical
7
value is ±1.645.

www.mathandstatistics.com

Step 5: Since the computed test statistic ' = −2.133 falls in the rejection region, reject the null
hypothesis. Conclude that at 0.10 level of significance, there is enough evidence that the percentage of
voters for the administration candidate is different from 65%.

Reference: Brase, C., (2012). Basic Statistics (5th Edition). Ortigas Avenue, Pasig City: Cengage
Learning

09 Handout 1 *Property of STI

Page 8 of 8