You are on page 1of 30

DMBA301: Research Methodology Manipal University Jaipur (MUJ)

MASTER OF BUSINESS ADMINISTRATION


SEMESTER 3

DMBA301
RESEARCH METHODOLOGY

Unit 10: Testing of Hypothesis 1


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

Unit 10
Testing of Hypothesis
Table of Contents

SL Fig No / Table SAQ /


Topic Page No
No / Graph Activity
1 Introduction - -
3
1.1 Learning Objectives - -
2 The Concept in Testing of Hypothesis 1 1
4–7
2.1 Types of Hypothesis 2 -
3 Steps in Testing of Hypothesis - 2 8 – 10

4 Test Statistic for Testing Hypothesis about 3 - 11 – 12


Population Mean
5 Tests Concerning Means - - 13 – 14
6 The Case of Single Population - 3 15 – 16

7 Tests for Difference Between Two - 4 17 – 19


Population Means
8 Test Concerning Population Proportion - 5 20
9 The Case of Single Population - 21

10 Hypothesis Testing for Difference between - 6 22 – 24


Proportions
11 Summary - - 25 – 26
12 Glossary - - 26 – 28
13 Terminal Questions - -
28 – 30
13.1 Answers - -
14 Suggested Books and References - - 30

Unit 10: Testing of Hypothesis 2


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

1. INTODUCTION

A hypothesis is, as we know, a supposition based on limited information. It is the reason why
research is conducted. Decision-makers regularly face situations wherein they are interested
in testing a hypothesis based on available information. They need to decide based on that
testing.

In social sciences, where direct information of population parameter(s) is rare, hypothesis


testing is the frequently used strategy—for deciding whether or not a sample data support a
hypothesis.

What is a Hypothesis?

The Hypothesis simply means a presumption or some


STUDY NOTE
speculation tobe proved or disproved. For a researcher, it is a
formal question that is intended to be solved. Generally, a The Hypothesis is
usually studied as the
research hypothesis isa predictive statement, capable of being main instrument in
analyzed by scientific methods that relates an independent research Its primary
function is to test
variable to some dependentvariable. For example, consider the how two variables in
statement: The automobile X is performing as well as the a study might be
related.
automobile Y. This hypothesis can be objectively proven and
tested. Therefore, we can say that a hypothesis is an argument
that can be validated.

1.1 Learning Objectives


After studying this chapter, the student will be able to:
❖ Explain the concept of Testing a Hypothesis
❖ Write the steps of Testing a Hypothesis
❖ Discuss the testing hypothesis of a single population
❖ Elucidate testing the population proportion of single and two population proportion

Unit 10: Testing of Hypothesis 3


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

ocess to test a hypothesis


2. CONCEPT comprises
IN TESTING three phases:
OF HYPOTHESIS
• Formulating a hypothesis
• Collecting appropriate evidence
• Analyzing evidence to conclude the validity

Figure 10.1 indicates this process diagrammatically. It is only after analyzing the evidence
that it can be concluded whether the hypothesis was true or false. For concluding about a
hypothesis, a statement is made whether the hypothesis is true or ‘the hypothesis is false’. A
hypothesis has to be clear, specific and testable. The study design, sample, data collection
method(s), data analysis, and conclusions needs to be free of any bias.

PHASE I PHASE I PHASE III

ANALYSE DATA TO DRAW


CONCLUSIONS ABOUT THE
FORMULATE COLLECT THE HYPOTHESIS --WHETHER
YOUR REQUIRED TRUE OR FALSE
ASSUMPTION DATA

Fig 1: The Process of Testing a


Hypothesis

2.1 Types of Hypotheses


There are six types of hypotheses, these are as follows:
STUDY NOTE
• Simple Hypothesis
The process to test a
hypothesis comprises three
It indicates a relationship between one dependent variable phases:

and a single independent variable. For example, eat more Formulating a hypothesis
Collecting appropriate
vegetables, you will be healthier. Here, eating more
evidence Analyzing
vegetables is an independent variable, while healthier is a evidence to conclude its
validity
dependent variable.

For Example: Regular consumption sugary drinks or carbonated drinks leads obesity

Unit 10: Testing of Hypothesis 4


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

• Complex Hypothesis

It indicates the relationship between two or more dependent and independent variables.
Eating more fruits and vegetables leads to glowing skin, weight loss, reducesthe risk of many
diseases, such as diabetes, heart disease, and some cancers.

For example:

Adults who 1) consume sugary beverages on a regular basis and 2. Genetic health issues are
more likely to 1) become obesity and 2) develop health issues or diabetics.

• Directional Hypothesis

A Directional Hypothesis is a prediction made by a researcher regarding a positive or


negative change, relationship, or difference between two variables of a population. The
relationship between the variables can predict. For example, children aged five years, eating
proper food over a five-year period, have higher IQ levels than childrennot having a proper
meal. This reveals the effect and direction of the effect.

For example The higher number of WhatsApp friends, the lower the life satisfaction score.

Fig 2: Types of Hypothesis

Unit 10: Testing of Hypothesis 5


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

• Non-Directional Hypothesis
Here there is no theory involved. It is a statement that a relationship occurs between two
variables, without predicting the correct nature (direction) of the relationship.

For Example Eating smarties will have a relationship effect on a student’s singing ability.

• Null Hypothesis
It presents a statement that is reverse to the hypothesis. It is a negative statement, meaning
there is no relationship between the dependent and independent variables. The symbol used
to express this statement is HO.

For Example: Students in the College score an average of 80 out of 100 in exams.

• Associative and Causal Hypothesis


Associative hypothesis arises when there is a variation in one variable resulting ina change
in the other variable. Here, the causal hypothesis proposes a cause-and-effect relationship
between two or more variables.

Unit 10: Testing of Hypothesis 6


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

SELF ASSESSMENT QUESTIONS – 1

1. _________ indicates the relationship between two or more dependent


and independent variables.
a) Simple hypothesis
b) Complex hypothesis
c) Directional hypothesis
d) Non-Directional hypothesis
2. ___________is a negative statement, meaning there is no relationship between
thedependent and independent variables.
a) Simple hypothesis
b) Complex hypothesis
c) Directional hypothesis
d) Non-Directional hypothesis
3. ___________is a statement that a relationship occurs between two variables,
withoutpredicting the correct nature of the relationship.
a) Simple hypothesis
b) Complex Hypothesis
c) Directional hypothesis
d) Non-Directional hypothesis

Unit 10: Testing of Hypothesis 7


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

3. STEPS IN TESTING A HYPOTHESIS


To test a hypothesis means to state—based on the data the
investigator has collected— whether or not the hypothesis looks STUDY NOTE

to be valid. In hypothesis testing, the main question is: whether Steps for hypothesis
testing refer to all those
to accept the null hypothesis or not to accept the null hypothesis. procedures that are
undertaken for selecting
Steps for hypothesis testing refers to all those procedures that between the two actions –
that is, acceptance or
are undertaken for making a selection between the two actions—
rejection of a null
acceptance or rejection of a null hypothesis. The various steps hypothesis.

involved in hypothesis testing are given below:

Step 1. Making a Formal Statement: This step comprises making a formal statementof
the null hypothesis (H0) and also of the alternative hypothesis (H1). This means that both the
hypotheses should be clearly specified, seeing the nature of the research problem. For
example, the average score in an aptitude test at the national level is 75. To assess a state’s
education system, the average score of 100 of the state’s students selected on random basis is

The state wants to know if there is any significant difference between the local scores andthe
national scores. The hypotheses may be stated as under:

Null hypothesis H0 : μ = 75

Alternative Hypothesis Ha : μ ≠ 75

The formulation of hypotheses is an essential step that has to be completed carefully


according to the object and nature of the problem. It also denotes whether to use a one-tailed
test or a value (or of the type lesser than a certain value), use a one-tailed test, but when Ha is
of the type `whether greater or smaller’ than a certain value, then use a two-tailed test.

Step 2. Selecting a Significance Level: The hypotheses are tested on a pre-determined level
of significance and the same can be specified. Normally in practice, either a 5% level or 1% level
is accepted for the purpose. The factors involved in selecting the level of significance are:

• The level of the difference between sample means


• The sample size

Unit 10: Testing of Hypothesis 8


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

• The adaptability of measurements within samples


• Whether the hypothesis is directional or non-directional

In short, the level of significance should be adequate with respect to the nature of enquiry.

Step 3. Deciding the Distribution to Use: The next step in hypothesis testing is to define the
correct sampling distribution. The choice usually remains between normal distribution and
the t-distribution.

Step 4. Selecting a Sample and Computing Value: The next step is to select random
sample(s) and calculate an appropriate value from the sample data about the test statistic
utilizing the relevant distribution. In other words, draw a sample to supply an empirical data.

Step 5. Calculating Probability: Calculate the probability that the sample effect would
diverge as widely as it has from expectations if the null hypothesis were true.

Step 6. Comparing Probability: Another step comprises comparing the probability thus
computed with the specific value for α, the significance level. If the determined probabilityis
equal to or smaller than a value—in the case of a one-tailed test and α/2 in the case of a two-
tailed test—then reject the null hypothesis. That is, accept the alternative hypothesis.
Whereas if the calculated probability is greater, then accept the null hypothesis. In case the null
hypothesis H0, is rejected, then the researcher runs the risk of (at most the level of
significance)Type I error. Whereas, if the null hypothesis H0, is accepted, then the researcher
runs some risk of committing Type II error.

Unit 10: Testing of Hypothesis 9


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

SELF ASSESSMENT QUESTIONS – 2

4. The null hypothesis is (H1) and alternative hypothesis is (H0). [True/False]


5. The hypotheses are tested on a pre-determined level of significance and the same
could be specified. [True/False]
6. A statement made about a population for testing a purpose is called, a .
a) Statistic
b) Hypothesis
c) Level of Significance
d) Test-Statistic

Unit 10: Testing of Hypothesis 10


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

4. TEST STATISTICS FOR TESTING HYPOTHESIS ABOUT POPULATION


MEAN

The Z-Test is a hypothesis test that helps to test a Population Mean, μ, besides a supposed
population mean, μ0. The z-test assumes normally distributed variables quantity or large
sample size. Then, the Central Limit Theorem assures a normally distributed sampling
distribution. In addition, σ, the standard deviation of the population needs to be known. In
real-life applications, this assumption is nearly never fulfilled, and thus, the z-test is rarely
applied. Though, it is a simple hypothesis test and thus, a helpful subject to start with.

The following steps are to be followed in testing a hypothesis about a population mean:

Step 1. Data: Define variable, sample size (n), sample mean (x), population standard deviation or

̅
sample standard deviation (s) if (𝑋) is unknown.

Step 2. Assumptions: The assumption is made in two sample cases as follows:

• Case1: Population is as normal, or approximately normally, distributed amongknown


or unknown variance (sample size n may be small or large).
• Case 2: Population is not normal with known or unknown variance n is greater than
30 (n is large that is, n≥30).

Step 3. Hypothesis: This can be:


Example 1:
Test that the population mean is different than 50

• Case I : H0:μ = μ0
H a: μ ≠ μ 0

Example 2: Test that the population mean is greater than 50

• Case II : H0: μ = μ0
H a: μ > μ 0

Unit 10: Testing of Hypothesis 11


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

Example 3: Test that the population mean is less than 50

• Case III : H0: μ = μ0


H a: μ < μ 0

WHEN THE VALUE OF SAMPLE SIZE (N)

POPULATION IS NORMAL OR NOT NORMAL POPULATION IS NORMAL

IS KNOWN IS NOT KNOWN IS KNOWN IS NOT KNOWN

Fig 3: Testing Hypothesis for Population


Mean μ

Unit 10: Testing of Hypothesis 12


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

5. TESTS CONCERNING MEANS


The Mean of a Population can be tested assuming different situations. Such as the population
may be normal or other than normal, it may be finite or infinite, the sample size may be large
orsmall, the variance of the population might be known or unknown, and the alternative
hypothesis might be two-sided or one-sided. The testing technique will differ in different
situations. It may consider some of the following important situations.

Situation 1: Population infinite, population normal, the sample size may be small or big
but variance of the population is known, Ha may be one-sided or two-sided:

In this situation, the z-test is used for the testing hypothesis of the mean and the test statistic
z is worked out as given:

𝑋̅ − 𝜇𝐻𝑜
𝑧=
𝜎𝑝 /√𝑛

Situation 2: Population finite, population normal, population finite, the sample size may
besmall or big but the variance of the population is known, Ha may be one-sided or
two-sided:

In this situation, the z-test is used and the test statistic z is worked out as given below (using
finitepopulation multiplier):

𝑋̅ − 𝜇𝐻𝑜
𝑧=
(𝜎𝑝 /√𝑛)𝑋√(𝑁 − 𝑛)/(𝑁 − 1)

Situation 3: Population infinite, population normal, sample size small, and variance of the
population unknown, Ha may be one-sided or two-sided:

In this situation, t-test is used and the test statistic t is worked out as given below:

𝑋̅ − 𝜇𝐻𝑜
𝑡=
𝜎𝑝 /√𝑛

Situation 4: Population finite, population normal, sample size small, and variance of the
population

Unit 10: Testing of Hypothesis 13


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

Not known, and Ha may be one-sided or two-sided:

In this situation, the t-test is used and the test statistic ‘t’ is worked out as given below
(using finite population multiplier):


s =

Situation 5: Population may not be normal but the sample size is high, the variance of
thepopulation may be unknown or known, and Ha may be one-sided or two-sided:

In this situation, use z-test and work out the test statistic z worked out as given below:

𝑋̅ − 𝜇𝐻𝑜
𝑡=
(𝜎𝑝 /√𝑛)𝑋√(𝑁 − 𝑛)/(𝑁 − 1)

Unit 10: Testing of Hypothesis 14


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

6. THE CASE OF A SINGLE POPULATION


A hypothesis test of a single population means, µ, using a
STUDY NOTE
Student’s t-distribution (often called a t-test), requires
In a hypothesis test of a single
important assumptions to be met for the test to work correctly. population mean µ using a
Student’s t-distribution,
The data required might be a simple random sample from a there are important
population that is approximately normally distributed. Apply assumptions that need to be
met for the test to work
the use of the sample standard deviation to estimate the correctly.

population standard deviation. Note that if the sample size is


greater than 30, a t-test will do, even if the population is not approximately normally-
distributed.

While performing a hypothesis test of a single population mean µ — using a normal distribution
(it is always called a z-test) – a researcher can take a simple random sample from the
population. The population tested is normally distributed or the sample size is greater than
30, or both. use the standard deviation.

While performing a hypothesis test of a single population


STUDY NOTE
proportion p, take a simple random sample from the
population. It must meet the conditions for a binomial The mean of the population
distribution, which are a particular number n of independent could be tested assuming
different situations. Such as the
trials. The results of any trial are success or failure, and each population may be normal or
other than normal, it may be
trial has a similar probability of a success p. The shape ofthe finite or infinite, the sample size
may be large or small, the
binomial distribution wants to be similar to the shape of the variance of the population
might be known or unknown
normal distribution. To make sure about this, the quantities
and the alternative hypothesis
np and nq must both be larger than five (np > 5 and nq > 5). might be two-sided or one-sided.

Then the binomial distribution of sample (estimated)


proportion could be approximated by the

normal distribution with µ = p and σ = . Remember thatq = 1 − p.

Unit 10: Testing of Hypothesis 15


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

Rare Events

Assume a property of the population (this assumption is the null hypothesis). Then collect
sample data randomly. If the sample has properties that might be unlikely to arise if the
assumption is true, then conclude that the assumption about the population is incorrect.
Remember that assumption is just an assumption—it is not a fact and it may or may not
be true. But the sample data is true and it is showing a fact that appears to contradict the
assumption.

SELF ASSESSMENT QUESTIONS – 3

7. For a random sample of 9 women, the average resting pulse rate is x = 76 beats
perminute, and the sample standard deviation is s = 5. The standard error of
the samplemean is___ .
a) 0.557
b) 0.745
c) 1.667
d) 2.778
8. The shape of the binomial distribution wants to be similar to the shape of the
normal distribution. [True/False]
9. A hypothesis test of a single population means µ using a Student’s
_________distribution.
a) t
b) z
c) binominal
d) simple random

Unit 10: Testing of Hypothesis 16


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

7. TESTS FOR DIFFERENCE BETWEEN TWO POPULATION MEANS


Often, in situations requiring decision-making, Need to understand, whether the factors of
two populations are alike or different. For example, whether female workers earn less than
male workers for the same job. Understand the technique of hypothesis testing for the
difference between population means. The null hypothesis for testing the difference between
means is normally stated as H0: μ1 = μ2, where μ1 is the population mean of one the

population and μ2 is population mean of the second population, if both the populations are
said to be normal populations. Alternative hypothesis might be:

• Ha: μ1 – μ2 < 0, which is the same as Ha: μ1 < μ2

• Ha: μ1 – μ2 > 0, which is the same as Ha: μ1 > μ2

• Ha: μ1 – μ2 ≠ 0, which is the same as Ha: μ1 ≠ μ2

Therefore, shall ascertain the acceptance or rejection areas for testing the hypotheses. There
might be different situations to assess the significance of the difference between the two
means,as follows:

Situation 1. Population variances are known or the samples occur to be large


samples:

In this situation, use the z-test for the difference in means and to solve or work out the
test statistic z, as given below:
𝑋̅1 − 𝑋̅2
𝑧=
2 2
𝜎𝑝1 𝜎𝑝2
√ +
𝑛1 𝑛2
In case σp1 and σp2 are not known, use σs and σ, respectively, in their places, calculating
2
∑(𝑋1𝑖 − 𝑋̅1 ) ∑(𝑋2𝑖 − 𝑋̅2 )
𝜎𝑠1 =√ 𝑎𝑛𝑑 𝜎𝑠1 = √
(𝑛1 − 1) (𝑛2 − 1)
2. Samples occur to be large but are assumed to have been taken from the same
population whose variance is known:

In this situation, use the z test for the difference in means and work out the test statistic z
given below:

Unit 10: Testing of Hypothesis 17


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

Case σp is not known, use 𝜎𝑠1.2 (combined standard deviation of the two samples) n its
place, to calculate

𝑛1 𝑋̅1 + 𝑛2 𝑋̅2
𝑋̅1.2 =
𝑛1 + 𝑛2

3. Samples occur to be lesser or small, samples and population variances not known
but assumed to be equal:

In this situation, use t-test for difference in means and work out the test statistic t as given
below:

Unit 10: Testing of Hypothesis 18


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

SELF ASSESSMENT QUESTIONS – 4

10. When testing the null hypothesis, using the confidence interval estimate of
the difference between two means, one would reject the null hypothesis when
_________________.
A. The higher confidence limit is less than zero.
B. The lower confidence limit is less than zero.
C. The confidence interval does not include zero.
D. The confidence interval includes zero.

11. Null hypothesis testing the difference between means is normally stated as
H0 : μ1 = μ2 , (True/False)

12. Where μ1 is population mean of one population and μ2 is population mean of


the second population, if both the populations to be normal populations.
(True/False)

Unit 10: Testing of Hypothesis 19


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

8. TEST CONCERNING POPULATION PROPORTION


In the case of qualitative phenomena, data -is based on the
presence or absence of an attribute(s). With such data,
STUDY NOTE
the sampling distribution may take the form of binomial The null hypothesis for
testing of difference
probability distribution whose mean would be equal to between means is normally
stated as H0 : μ1 = μ2 ,
n.p and standard. √𝑛. 𝑝. 𝑞 probability of success, q
represents the probability of failure, such that p + q = 1,
and n is the size ofthe sample. Taking the mean number of successes, and the standard
deviation of the number of successes, record the proportion of successes in each sample.
In whichever the case, the mean and standard deviation (or the standard error) of the
sampling distribution formula are as under:
Mean proportion of successes (n.p)/n = P and standard deviation of the proportion of
successes

If n is larger, the binomial distribution leans to become normal distribution, and as such, for
proportionately testing purposes, make use of the test statistic z as under:
√𝑝. 𝑞
𝑧=
𝑛
Where 𝑝̂ is the sample proportion.
For testing the proportion, formulate H0 and Ha and construct the rejection region. Assuming

normal approximation of the binomial distribution, for a pre-determined level of significance


and then judge the significance of the observed sample result.

SELF ASSESSMENT QUESTIONS – 5

13. The null hypothesis for testing of difference between means is normally stated as

H0 : μ1 ≠ μ2 . (True/False)

Unit 10: Testing of Hypothesis 20


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

9. THE CASE OF A SINGLE POPULATION


The One-Sample Proportion Test is used to compute whether a population proportion (P1)
is significantly different from a hypothesized value (P0). This is known as the hypothesis of
inequality. The hypotheses might be specified in terms of the proportions, their difference,
their ratio, or their odds ratio. However, all four hypotheses have the same outcome in the
same test statistics.

This procedure computes sample size and statistical power for analyzing a single proportion
using either the exact test or other approximate z-tests. The exact test results are
hypergeometric) based on calculations using the binomial (and distributions. Since the
analysis of many different test statistics is available, their statistical power might be compared
to find the most suitable test for a given situation.

This process has the capability of calculating power using both the normal estimates and
binomial records for all tests. Some sample size programs utilize only the normal
approximation to the binomial distribution for power and sample size assessments. The normal
approximation is exact for large sample sizes and proportions between 0.2 and 0.8, roughly.

Test Statistics

Many different test statistics have been suggested for testing a single proportion. Most of these
were suggested before computers or hand calculators were widely available. To make this
assessment easy, the power and significance of many tests of a single proportion are available.

Exact Test

The test statistic is r, the number of successes in n trials. This test should be the standard
against which other test statistics are judged. The significance level and power are computed
by enumerating the possible values of r, computing the probability of each value, and then
computing the corresponding value of the test statistic. Hence the values that are stated in the
output for these tests are exact, not approximate.

Unit 10: Testing of Hypothesis 21


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

10. HYPOTHESIS TESTING FOR DIFFERENCE BETWEEN PROPORTIONS


If two samples are obtained from different populations, researchers may be interested in
knowing whether the difference between the proportion of successes is significant or not
significant. In such a case, the hypothesis that the difference between the proportion of success

in sample one (p̂1) and the proportion of success in sample two (p̂2), is due to fluctuation or

variation of random sampling. In other words, take the null hypothesis as H0: p1 p2. For testing
the significance of difference, to find and check the test statistic as under:

Where p̂1 = proportion of success in sample one

p̂2 = proportion of success in sample two q̂ = 1 – p̂1


q̂2 = 1 – p̂2
n1 = size of sample one
n2 = size of sample two

= the standard error of the difference between two sample proportions.*

Construct the rejection region(s), depending upon the Ha for a given level of significance. On
its basis, judge the significance of the sample result for accepting or rejecting H0.

Sample Problem

A research department’s New Research Experimental Unit is testing two drugs newly
developed to decrease Covid-19. The drugs are directed to two different sets of animals. In
group one, 350 of 600 animals tested reacted to Drug 1 and in group two, 260 of 500 animals
tested to respond to Drug 2. The research unit wants to test whether there is a difference

Unit 10: Testing of Hypothesis 22


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

betweenthe efficacy of the said two drugs at 5 percent level of significance. How will you deal with
thisproblem?

Solution: The null hypothesis is that there is no significance difference between the two
drugs. That is,

H0: p̂1 = p̂2

The alternative hypothesis that is Ha can be selected as that there is a differencebetween

the drugs. That is,

Ha: p̂1 ≠ p̂2

and the given information can be stated as: p̂1 = 350/600 = 0.583
= 600
p̂2 = 260/500 = 0.520
q̂2 = 1 – p̂2 = 0.480
n2 = 500

As Ha is two-sided, determine the rejection regions by applying a two-tailed test at a 5% level,

whichcomes as under using the normal curve area table:

R : | z | > 1.96

The z observed value is 2.093, which is the rejection region. Thus, H0 or Null hypothesis is
rejected in favor of Ha and as such concluded that the difference between the efficacy of the

two drugs is significant.

Unit 10: Testing of Hypothesis 23


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

SELF ASSESSMENT QUESTIONS – 6

14. Some sample size programs utilize only the normal approximation to the
binomial distribution for power and sample size assessments. (True/False)
15. There are 100 students in a Tamil Nadu government school. In the four
states, inclusive of this school, the number of students is 2000. In a random
sample study, 20were found smokers in the school. The proportion of smokers in
the school is 0.05. Is there a significant difference between the proportion of
smokers in the school and in Tamil Nadu? Test at 5 % level.
A. Accept hypothesis
B. Reject the hypothesis

Unit 10: Testing of Hypothesis 24


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

11. SUMMARY

• The Hypothesis is usually studied as the main instrument in research. Its primary
functionis to suggest new tests and observations. Several experiments are taken out
with the deliberate object of testing hypotheses.
• To test a hypothesis, the process comprises three phases: One, Formulating a
hypothesis, Two, Collecting appropriate evidence, and three, Analyzing evidence to
concludeits validity.
• Simple Hypothesis indicates a relationship between one dependent variable and a single
independent variable.
• Complex Hypothesis indicates the relationship between two or more dependent and
independent variables.
• Non-Directional Hypothesis is a statement that a relationship occurs between two
variables, without predicting the correct nature (direction) of the relationship.
• Null Hypothesis is a negative statement, meaning there is no relationship between the
dependent and independent variables. The symbol is expressed as HO.

• Associative hypothesis arises when there is a variation in one variable resulting in a


• change in the other variable.
• Steps to be followed in Testing a Hypothesis about the population mean are data,
assumption and hypothesis.
• While performing a hypothesis test of a single population proportion p, take a simple
random sample from the population. It must meet the conditions for a binomial
distribution which is there are a particular number n of independent trials. The results
ofany trial are success or failure, and each trial has a similar probability of a success p.
• whether the factors of two populations are alike or different. For example, whether
female workers earn less than male workers for the same job. The null hypothesis for
testing of difference between means is normally stated: as H0 : μ1 = μ2 , where m1 is
population mean of one population and m2 is population mean of the second population,
• In the case of qualitative phenomena, data is based on the presence or absence of an
attribute(s). With such data, the sampling distribution may take the form of a binomial
probability distribution whose mean would be equal to n.p and standard.

Unit 10: Testing of Hypothesis 25


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

• If two samples are obtained from different populations, one may be interested in
knowing whether the difference between the proportion of successes is significant ornot
significant.

12. GLOSSARY
Causal Hypothesis: This is a statement hypothesizing that the independent variable impacts
the dependent variable in some way.

Central Tendency: This is a way of defining or characterizing typical, average, or common


values in certain distribution.

Confidence interval estimate: A confidence interval, in statistics, refers to the probability


that a population parameter will fall between a set of values for a certain proportion of times.

Chi-Square Analysis: This is a normal non-parametric statistical test which relates an


expected proportion or ratio to a definite proportion or ratio.

Construct Validity: Indicates an agreement between a theoretical concept and a specific


measuring device, such as observation.

Data: This is the factual evidence—as measurements or statistics—used as a basis for


discussion, reasoning, or calculation.

Dependent Variable: This is a variable that varies, at least in part, due to the effect of the
independent variable. In other words, its value or estimate `depends’ on the value of the
independent variable.

Group Behavior: Refers to the behaviors of a group as a unit, as well as the behavior of an
individual, as induced by his or her membership in a group.

Hypothesis: This is a tentative description based on theory to expect a causal relationship


between variables.

Unit 10: Testing of Hypothesis 26


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

Hypergeometric: The probability density function is the hypergeometric. This is the most
basic one because it is created by combining our knowledge of probabilities from Venn
diagrams, the addition and multiplication rules, and the combinatorial counting formula.

Internal Consistency: This is the area to which all questions or items measure the same
characteristic, skill, or quality.

Null Hypothesis: It is the proposition, to be assessed statistically that the experimental


intervention has no effect,’ meaning that the action and control groups will not differ as an
outcome of the intervention.

One tailed test: A one-tailed test in hypothesis testing is a test defined so that it matchesa
one-sided hypothesis. The “tail” references one half of the frequency distribution of the
statistic one is using (e.g. a z score or a t score).

Participant: Individuals whose physiological and/or behavioral qualities and responses are
the objects of study in a research project.

Population: This is the target group under investigation.

Probability: Refers to the chance that an event will occur randomly. As a statistical measure,
it is expressed as p.

Random Sampling: Random sampling can be completed by first numbering the population,
then choosing the sample according to a table of random numbers or using a random-number
computer generator.

Rigor: The degree to which research methods are carefully and accurately carried out in order
to recognize important influences occurring in an experimental study.

Rejection regions: A rejection region (a.k.a. critical region) in a Null Hypothesis Statistical
Test is a part of the parameter space such that observing a result that falls under it will lead to
the rejection of the null hypothesis.

Sample: It is the population studied in a particular study.

Unit 10: Testing of Hypothesis 27


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

Sampling Error: This is the degree to which the results from the sample diverge from those
that would be acquired from the entire population.

Standard Deviation: Is a measure of variation that shows the typical distance among the
scores of a distribution and the mean.

Statistical Analysis: Is the application of statistical processes and theory to the compiling,
presentation, analysis, and interpretation of numerical data.

Type II Error: Refers to the failure of an investigator to reject a null hypothesis that is actually
false. This is also called beta error.

13. TERMINAL QUESTIONS


SHORT ANSWER QUESTIONS

Q1. Define Hypothesis.

Q2. Write the three process of hypothesis.

LONG ANSWER QUESTIONS

Q1. Discuss the steps involved in hypothesis testing.

Q2. Explain tests for difference between two population means.

13.1 ANSWERS

SELF ASSESSMENT QUESTIONS

1. Simple hypothesis
2. Non-Directional hypothesis
3. Directional hypothesis
4. False
5. True
6. Hypothesis
7. 1.667
8. True

Unit 10: Testing of Hypothesis 28


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

9. Binomial distribution
10. the confidence interval does not include zero
11. 38
12. True
13. False
14. True
15. Reject

TERMINAL QUESTIONS

SHORT ANSWER QUESTIONS

Answer 1: Hypothesis is defined as a proposition or a set of proposition set forth as an


account of the occurrence of some stated group of phenomena—either asserted purely as a
provisional conjecture to guide some investigation, or accepted as highly possible in the light
of ascertained facts.

Answer 2: The three phases in the process of making a hypothesis are: One, Formulating a
hypothesis; two, Collecting appropriate evidence; and three, analyzing evidence to conclude
its validity.

LONG ANSWER QUESTIONS

Answer 1: The steps involved in hypothesis testing are:

Step 1. Making a formal statement Step 2. Selecting a significance level Step 3. Deciding the
distribution to use

Step 4. Selecting a random sample and computing an appropriate valueStep 5. Calculation of


the probability 6.Comparing the probability Answer 2

The null hypothesis for testing the difference between means is normally stated as
H0 : μ1 = μ2 , where μ1 is population mean of one population and μ2 is population mean

Unit 10: Testing of Hypothesis 29


DMBA301: Research Methodology Manipal University Jaipur (MUJ)

of the second population, if both the populations are said to be normal populations.
Alternative

hypothesis might be:

Ha: μ1 – μ2 < 0, which is the same as Ha: μ1 < μ2

Ha: μ1 – μ2 > 0, which is the same as Ha: μ1 > μ2

Ha: μ1 – μ2 ≠ 0, which is the same as Ha: μ1 ≠ μ2

Therefore, shall ascertain the acceptance or rejection areas for testing the hypotheses. There
might be different situations to assess the significance of difference between the two means,
as follows:

Situation 1. Population variances are known or the samples occur to be large samples:

14. SUGGESTED BOOKS AND REFERENCES


BOOKS

• C. R. Kothari, (2007), Research Methodology: Methods and Techniques: New Age


International (P) Ltd, New Delhi.
• Mukul Gupta and Deepa Gupta, (2009), Research Methodology: PHI Publication, India
• R. Panneerselvam, (2014), Research Methodology: PHI Learning Private Ltd, New Delhi.

REFERENCES

• https://sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs704_hypothesistest-means-
proportions/bs704_hypothesistest-means-proportions_print.html
• https://www.sagepub.com/sites/default/files/upm-binaries/40007_Chapter8.pdf
• https://stats.libretexts.org/Bookshelves/Introductory_Statistics/Book%3A_
Introductory_Statistics_(Shafer_and_Zhang)/08%3A_Testing_Hypotheses/8.01%3A_
The_Elements_of_Hypothesis_Testing
• https://opentextbc.ca/introductorybusinessstatistics/chapter/hypothesis-testing-2/
• https://courses.lumenlearning.com/wmopen -concepts-statistics/chapter/
hypothesis-test-for-a-difference-in-two-population-means-1-of-2/

Unit 10: Testing of Hypothesis 30

You might also like