Statistics and Probability

Test of Hypotheses

Nida C. Flores
Teacher III
 Statistical Hypothesis Testing

Overview

The concept of statistical hypothesis testing is considered to be the most important area of
statistical inference. This contributes to the formulating of rules that would lead us to
deciding whether or not a particular statement or claim about a population should be
rejected. This chapter focuses on testing hypotheses concerning population mean and
population proportion- how to formulate competing hypotheses, how to identify the level of
significance and the appropriate rejection region, and how to determine which test statistic
to use in testing the hypotheses
 Statistical Hypothesis Testing

LESSON 1
Statistical Hypothesis Testing
 Statistical Hypothesis Testing

Case Study

A particular company claims that more than65percent of its employees earn more than Php216,000 annually.
How can this claim be validated without taking a census of the whole employee population?
What evidence is needed to test this claim?
 Statistical Hypothesis Testing

Lesson Objectives
 Illustrates : (a) null hypothesis;(b) alternative hypothesis;(c) level of
significance; (d) rejection region; and (e) types of errors in hypothesis
testing
(M11/12SP- IVa-1)
 Identifies the parameter to be tested given a real-life problem(M11/12SP-

IVa-3)
 Statistical Hypothesis Testing

Expected Outcomes
 Illustrate null hypothesis, level of significance, rejection region, and types of
errors in hypothesis testing;
 Calculate the probabilities of committing a Type I and Type II errors
 Identify the parameter to be tested given a real-life problem
 Statistical Hypothesis Testing

Statistical Hypothesis
We estimated the value/s of some parameters. In some cases, we can also use sample data to make
certain decisions on some claims about a population parameter. If we obtain evidence from the
sample that is inconsistent with the given hypothesis, we decide to reject this hypothesis. If we have
insufficient evidence to show otherwise, we do not reject the hypothesis.

Example.
a. A researcher can determine whether the children in a particular shelter has an attention span that
is less than 2 hours
b. A surveyor can check whether or not 45 percent of the commercial establishments in an area are a
safe distance away from the earthquake danger zone.

In example (a) the researcher is concerned with the value of the mean , while the surveyor,
deals with proportion
 Statistical Hypothesis Testing

STATISTICAL HYPOTHESIS
A statistical hypothesis is a claim or assertion about a parameter or some parameters

HYPOTHESIS:
1. Let be the mean number of text messages from mobile phones.
= 87 ( The mean number of text messages from mobile phones is 87)

Let the proportion of students that do not attend the general assembly
0.15 ( the proportion of students that do not attend the general assembly is 0.15)

= 87, the other hypothesis could be 87.

,
 Statistical Hypothesis Testing

HYPOTHESIS TESTING
A test of hypotheses is a method to test the claims about a characteristic of a population through the use
of sample data. This is also called a test procedure

NULL AND ALTERNATIVE HYPOTHESIS FORMULATION

To form competing claims about a parameter , start with assuming that a particular hypothesis is
the correct claim, called null hypothesis. If there is strong evidence against the null hypothesis, we reject it in
favor of the competing claim called alternative hypothesis. They are always formulated together to decide
when to reject the null hypothesis.

The null hypothesis, , is a claim about a characteristic of the population that is initially assumed as a true
statement.
The alternative hypothesis, , is a competing claim or statement to the null hypothesis.

We set up the null hypothesis with the hope that it will be rejected. We reject in favor of only
if the evidence based on the sample data strongly implies that is a false statement. will not be rejected
if the sample data does not provide strong evidence, thus, we fail to reject. When we do test of hypotheses,
either we reject or we fail to reject
 Statistical Hypothesis Testing

The null hypothesis is of this form.

The alternative hypothesis will be one of the following.

This further explains that the null hypothesis always indicates an exact hypothesized value of the
parameter. The alternative hypothesis specifies that the population parameter is less than, greater
than, or not equal to the hypothesized value in the null hypothesis.

will always be stated as = (population characteristic =hypothesized value,

Sample Problem
The owner of a factory that sells a particular bottled fruit juice claims
that the average capacity of a bottle of their product is 250 ml. To test
the claim, a consumer group get 100 samples of such bottles and
calculate the capacity of each bottle. The observed mean capacity of
the 100 bottles is 243ml. The standard deviation is 10 ml. In the
example, the owner‘s statement (called claim) is a general statement.
The claim is that the capacity of all their bottled products is250 ml. On
the other hand, the consumer group has a sample which is = 243ml.
clearly a sample mean. There is a difference of 7 m
The two hypotheses would be;
The bottled drinks contain 250 ml. per bottle( This is the claim)
• : The bottled drink do not contain 250 ml per bottle.( This is the
opposite of the claim)
• In symbols, = 250 and : =

So, the appropriate alternative hypothesis is , :

 Statistical Hypothesis Testing

Activity
Write the null and alternative hypotheses for each situation
1.A manufacturer of car tires wants to test whether their machine still
produces car tires with a mean diameter of 16 inches.

2.An airline company claims that the percentage of their flights prone
to delays weekly due to air traffic and other technical issues does not
exceed 28 percent of their total flights
 Statistical Hypothesis Testing

ANSWER TO ACTIVITY

1. Answer: Let be the mean diameter of car tires produced by the manufacturer.
: = 16
: 16
We choose 16 as the alternative hypotheses. This is our competing claim that the
machine does not produce car tires with a mean diameter equal to 16.

2. Answer: Let p be the percentage of flights prone to delay

: = 0.28
:
We choose : ,since we want to reject the null hypothesis that the percentage of flights
prone to delay is greater than or equal to 28 percent.
 Statistical Hypothesis Testing

REJECTION OR CRITICAL REGION

A test procedure can only show strong evidence for the alternative hypothesis by rejecting
the null hypothesis. But when the null hypothesis is not rejected , this implies that there is lack of
strong support against this claim.
A test procedure is specified in terms of a particular test statistic and rejection rule.
Most of the time, the test statistic used to test hypotheses about a parameter is a point estimator of
the of If a test statistic takes on a value that is deemed very unlikely when the null hypothesis is true,
then the null hypothesis can be rejected. The set or collection of these unlikely values for the test statistic
is called the rejection or critical region
 Statistical Hypothesis Testing

ERRORS IN TEST OF HYPOTHESES

If the null hypothesis is true but is rejected, the decision is incorrect. This is the Type I error.
If the null hypothesis is false but fails to be rejected, the decision is also incorrect. This is called
a Type II error

A Type I error is the error of rejecting the null hypothesis when it is true.
A Type II error is the error of failing to reject the null hypothesis when it is false

We ca n only guarantee an error free conclusion if we base our decision on a census of the population.
There will always be risk of error when we base the decision on a sample data.
 Statistical Hypothesis Testing

Examples.
1.A manufacturer of car tires wants to test whether their machine still produces car tires with a mean diameter of 16 inches.
• : = 16
• : 16
Type I error : Believing that the machine produces car tires with non- identical diameters when it actually produces identical
ones
Type II error: Accepting the claim that the machine produces car tires with a diameter of 16 inches when it does not
2.An airline company claims that the percentage of their flights prone to delays weekly due to air traffic and other technical
issues does not exceed 28 percent of their total flights.
• : = 0.28
• :
Type I error: To believe the that the airline has a percentage of delayed flight that ismorethan28 percent when in reality, it has
less.
Type II error : To accept the airline’s claim that the percentage of their delayed flights is less than28 percent weekly, when it
is more than this in actual.
 Statistical Hypothesis Testing

LEVEL OF SIGNIFICANCE
The probability of a Type I error is denoted by .This is also called the level of significance of the test.
While, the probability of a type II error is denoted by .
The standard test procedure allows researchers to control the value of .If we assume
to be 5 percent,(, this means the probability of rejecting a true null hypothesis is 5 percent.
This also results to a test procedure with a 5 % probability of Type I error. Choosing a small implies
that the researcher wants to have a test procedure for which the possibility of a Type I error is quite small.
The standard test procedure does not give the researcher direct control over
I error increases the risk of a Type II error. It is important that the researcher decides
on the probabilities of these types of errors by considering the consequences of committing the error
 Statistical Hypothesis Testing

SPECIFYING TEST PROCEDURE:

1. Assess the consequences of the Type I and Type II errors in the given situation.
2. Determine the greatest value of that is tolerable in the situation
3. Use a test procedure that applies this maximum value of.
4. Note, that using a smaller increases the value of ,so it is advised not to use a smaller
than it has to be.

Example:
An air- conditioner manufacturer claims that 85 percent of its home air- conditioning units
will not need after-sale service during the first year of use. Suppose this claimis to be tested by a consumer
group by taking a random sample of the buyer’s units.
The manufacturer will not admit false advertising unless the sample evidence strongly shows that
the proportion of the air conditioning units without need for repair or service is less than 85 percent(p<0.85)
 Statistical Hypothesis Testing
SPECIFYING TEST PROCEDURE:
Example:
An air- conditioner manufacturer claims that 85 percent of its home air- conditioning units
will not need after-sale service during the first year of use. Suppose this claimis to be tested by a consumer
group by taking a random sample of the buyer’s units.
The manufacturer will not admit false advertising unless the sample evidence strongly shows that
the proportion of the air conditioning units without need for repair or service is less than 85 percent(p<0.85)
Hypotheses:
: = 0.85
: 0.85
Type I error : To believe that the manufacture’s claim that 85 % of its home air conditioning units will not need
after-sales service, when this is not the case.
Interpretation: This is false advertising and the possible consequence is that consumers are made to believe
this claim and will continue to patronize the product.
Type II error: This is to conclude that less than 85 percent of its home air conditioning units will not need
after-sale service, when the manufacturer is actually telling the truth
Interpretation: A possible consequence of this type of error is the decline of purchases of the air-conditioning
units from the manufacturer due to negative consumer impression

Which do you think should be used as the value of I n the test procedure, = 0.01 or = 0.1 ?
DIRECTIONAL AND NON- DIRECTIONAL HYPOTHESIS
• When the alternative hypothesis utilizes the symbol, the test is said to be non- directional.
• When the alternative hypothesis utilizes the or the symbol, the test is said to be directional
• A non- directional test is also called a two-tailed test.
• A directional test may either be left- tailed or right tailed.

In the formulation of alternative hypothesis words like greater, efficient, improves, effective,
increases and so on, suggest a right – tailed direction while words like decrease, less than , smaller and
the like, suggest a left- tailed direction
Graphical Representations of the two-tailed test
Two-tailed Test
H0: =  is divided equally between
the two tails of the critical
H1:  region

Means less than or greater than

Copyright © 2007 Pearson Education, Inc Publishing as Pearson Addison-Wesley.

Slide 35

Non- directional (Two-tailed).

The Probability is found on both tails of the distribution
Graphical Representations of the right-tailed test
Right-tailed Test
H0: =
H1: >
Points Right

Copyright © 2007 Pearson Education, Inc Publishing as Pearson Addison-Wesley.

Slide 36

Directional(One- tailed, right tail)

The probability is found at the right tail of the distribution
Graphical Representations of the left -tailed test
Left-tailed Test
H0: =
H1: <
Points Left

Copyright © 2007 Pearson Education, Inc Publishing as Pearson Addison-Wesley.

Slide 37

(One –tailed, left tail)

The probability is found at the left tail of the distribution.
 Statistical Hypothesis Testing
Your Turn!
A recent survey claims that82% of the residents in NCR favor the proposed tax exemption policy .
Suppose this claim is to be tested by taking a random sample of residents and asksthem of their opinion
about the new tax policy. State the null and alternative hypotheses and the Type Iand Type Iierrors.
Determine the appropriate level of significance to use.
SEATWORK:
State the null and the alternative hypothesis in words and in symbols.
State whether the test is directional or non- directional.
1. A teacher wants to know if listening to popular music affects the performance of pupils. A
class of 50 grade 11 students was used in the experiment. The mean score was 83 and the
standard deviation is 5.0. A previous study revealed that =82 and the standard deviation =
10.
2. A company that produces a particular brand of snacks claims that each box of snack pack
they produce has a mean weight of 15 grams. Consumers would not want to purchase an
item with less than the advertised weight.
SOLUTION:

: = 82 and : 82, = mean score

b. There is no clue as to the direction of the investigation . The phrase affects performance
implies either an increase or a decrease in performance. So the test is non directional.
2. Let the mean weight of snack pack.
• : = 15 and : 15
• Choosing : 15 means the competing null hypothesis that the snack packs produced have a
mean weight of 15 grams or more
REAL –LIFE APPLICATION
A farmer believes that using organic fertilizers on his plants will yield greater
income. His average income from the past was P200, 000.00 per year. State the
hypotheses in symbols.
SOLUTION
• : = 200,000.00
• The phrase “ greater income” is associated with the greater than direction
• So, : 200,000.00
WHAT HAVE YOU LEARNED?

• What is hypothesis testing

• What are the types of hypothesis?
• When do we say that a test is directional or non-directional? right-
tailed or left- tailed?
QUIZ

State the null hypothesis and the alternative hypothesis. Sketch the graph of each item.
1.Are less than 45% of the Grade 11 students are left handed?
2. Is the mean number of students applying for scholarship different from 300?
3. Is the percentage of voters registered through biometrics more than 90%?
4. Is there strong evidence that more than 20 percent of the traffic enforcers are
deployed in the streets on time?
5.A sales agent claims that the average market value of the houses they sell is at least
Php 1.5 million.
Activity
Use the Internet to collect data worth investigating
from government agencies , ex. Mean age of minors
committing crimes, mean percentage of grade 7
dropouts, mean percentage of accidents caused by
motorcycles). Formulate the null hypothesis and the
alternative hypothesis for your data.