Types Of Research Design:

There are different types of research designs. They may be broadly categorized as:

(1) Exploratory Research Design;

(2) Descriptive and Diagnostic Research Design; and
(3) Hypothesis-Testing Research Design.

1. Exploratory Research Design:

The Exploratory Research Design is known as formulative research design. The main objective of using
such a research design is to formulate a research problem for an in-depth or more precise investigation,
or for developing a working hypothesis from an operational aspect. The major purpose of such studies is
the discovery of ideas and insights. Therefore, such a research design suitable for such a study should be
flexible enough to provide opportunity for considering different dimensions of the problem under study.
Usually, the following three methods are considered in the context of a research design for such studies.
They are (a) a survey of related literature; (b) experience survey; and (c) analysis of insight-stimulating
instances.

2. Descriptive And Diagnostic Research Design:

A Descriptive Research Design is concerned with describing the characteristics of a particular individual
or a group. Meanwhile, a diagnostic research design determines the frequency with which a variable
occurs or its relationship with another variable. In other words, the study analyzing whether a certain
variable is associated with another comprises a diagnostic research study. On the other hand, a study
that is concerned with specific predictions or with the narration of facts and characteristics related to an
individual, group or situation, are instances of descriptive research studies. Generally, most of the social
research design falls under this category.

The research design in such studies should be rigid and not flexible. Besides, it must also focus attention
on the following:
a) Formulation of the objectives of the study,
b) Proper designing of the methods of data collection,
c) Sample selection,
d) Data collection,
e) Processing and analysis of the collected data, and
f) Reporting the findings.

3. Hypothesis-Testing Research Design:

Hypothesis-Testing Research Designs are those in which the researcher tests the hypothesis of causal
relationship between two or more variables. These studies require procedures that would not only
decrease bias and enhance reliability, but also facilitate deriving inferences about the causality.
Generally, experiments satisfy such requirements. Hence, when research design is discussed in such
studies, it often refers to the design of experiments.
Hypothesis:

Hypothesis may be defined as a proposition or a set of propositions set forth as an explanation for
the occurrence of some specified group of phenomena either asserted merely as a provisional
conjecture to guide some investigation in the light of established facts (Kothari, 1988). A research
hypothesis is quite often a predictive statement, which is capable of being tested using scientific
methods that involve an independent and some dependent variables.

i. Students who take tuitions perform better than the others who do not receive tuitions or,
ii. The female students perform as well as the male students.

These two statements are hypotheses that can be objectively verified and tested. Thus, they
indicate that a hypothesis states what one is looking for. Besides, it is a proposition that can be put
to test in order to examine its validity.

Characteristics Of Hypothesis:

hypothesis should have the following characteristic features:-

i. A hypothesis must be precise and clear. If it is not precise and clear, then the inferences drawn on its
basis would not be reliable.

ii. A hypothesis must be capable of being put to test. Quite often, the research programmes fail owing to
its incapability of being subject to testing for validity. Therefore, some prior study may be conducted by
the researcher in order to make a hypothesis testable. A hypothesis is tested if other deductions can be
made from it, which in turn can be confirmed or disproved by observation (Kothari, 1988).

iii. A hypothesis must state relationship between two variables, in the case of relational hypotheses.

iv. A hypothesis must be specific and limited in scope. This is because a simpler hypothesis generally
would be easier to test for the researcher. And therefore, he/she must formulate such hypotheses.

v. As far as possible, a hypothesis must be stated in the simplest language, so as to make it understood
by all concerned. However, it should be noted that simplicity of a hypothesis is not related to its
significance.

vi. A hypothesis must be consistent and derived from the most known facts. In other words, it should be
consistent with a substantial body of established facts. That is, it must be in the form of a statement
which is most likely to occur.

vii. A hypothesis must be amenable to testing within a stipulated or reasonable period of time. No
matter how excellent a hypothesis, a researcher should not use it if it cannot be tested within a given
period of time, as no one can afford to spend a life-time on collecting data to test it.
Concepts Relating To Testing Of Hypotheses

Testing of hypotheses requires a researcher to be familiar with various concepts concerned with it such
as:
1) Null Hypothesis And Alternative Hypothesis:
In the context of statistical analysis, hypotheses are of two types viz., null hypothesis and alternative
hypothesis. When two methods A and B are compared on their relative superiority, and it is assumed
that both the methods are equally good, then such a statement is called as the null hypothesis. On the
other hand, if method A is considered relatively superior to method B, or vice-versa, then such a
statement is known as an alternative hypothesis. The null hypothesis is expressed as H0, while the
alternative hypothesis is expressed as Ha. For example, if a researcher wants to test the hypothesis that
the population mean () is equal to the hypothesized mean (H0) = 100, then the null hypothesis should
be stated as the population mean is equal to the hypothesized mean 100. Symbolically it may be written
as:-
H0: = = H0 = 100
If sample results do not support this null hypothesis, then it should be concluded that something else is
true. The conclusion of rejecting the null hypothesis is called as alternative hypothesis H1. To put it in
simple words, the set of alternatives to the null hypothesis is termed as the alternative hypothesis. If H0
is accepted, then it implies that Ha is being rejected. On the other hand, if H0 is rejected, it means that
Ha is being accepted. For H0: = H0 = 100, the following three possible alternative hypotheses may be
considered:

2) The Level Of Significance:

In the context of hypothesis testing, the level of significance is a very important concept. It is a
certain percentage that should be chosen with great care, reason and insight. If for instance, the
significance level is taken at 5 per cent, then it means that H0 would be rejected when the sampling
result has a less than 0.05 probability of occurrence when H0 is true. In other words, the five per
cent level of significance implies that the researcher is willing to take a risk of five per cent of
rejecting the null hypothesis, when (H0) is actually true. In sum, the significance level reflects the
maximum value of the probability of rejecting H0 when it is actually true, and which is usually
determined prior to testing the hypothesis.
3) Test Of Hypothesis Or Decision Rule:
Suppose the given hypothesis is H0 and the alternative hypothesis H1, then the researcher has to
make a rule known as the decision rule. According to the decision rule, the researcher accepts or
rejects H0. For example, if the H0 is that certain students are good against the H1 that all the students
are good, then the researcher should decide the number of items to be tested and the criteria on
the basis of which to accept or reject the hypothesis.
4) Type I And Type II Errors:
As regards the testing of hypotheses, a researcher can make basically two types of errors. He/she
may reject H0 when it is true, or accept H0 when it is not true. The former is called as Type I error and
the latter is known as Type II error. In other words, Type I error implies the rejection of a hypothesis
when it must have been accepted, while Type II error implies the acceptance of a hypothesis which
must have been rejected. Type I error is denoted by (alpha) and is known as error, while Type II
error is usually denoted by (beta) and is known as error.
5) One-Tailed And Two-Tailed Tests:
These two types of tests are very important in the context of hypothesis testing. A two-tailed test
rejects the null hypothesis, when the sample mean is significantly greater or lower than the
hypothesized value of the mean of the population. Such a test is suitable when the null hypothesis is
some specified value, the alternative hypothesis is a value that is not equal to the specified value of
the null hypothesis.
Data Collection & Sources of Data

1. Lesson Outline:
2. Primary Data, Secondary Data
3. Investigation
4. Indirect Oral Methods Of Collecting Primary Data
5. Direct Personal Interviews
6. Information Received Through Local Agencies
7. Mailed Questionnaire Method
8. Schedules Sent Through Enumerators

It is important for a researcher to know the sources of data which he requires for different
purposes. Data are nothing but the information. There are two sources of information or
data they are - Primary and Secondary data. The data are name after the source. Primary
data refers to the data collected for the first time, whereas secondary data refers to the data
that have already been collected and used earlier by somebody or some agency. The
selection of a particular source depends upon the
1. Purpose and scope of enquiry,
2. Availability of time,
3. Availability of finance
4. Accuracy required,
5. Statistical tools to be used,
6. Sources of information (data), and
7. Method of data collection.

Methods of Collecting Primary Data:

Primary data may be obtained by applying any of the following methods:

1. Direct Personal Interviews.
2. Indirect Oral Interviews.
3. Information from Correspondents.
4. Mailed Questionnaire Methods.
5. Schedule Sent Through Enumerators.
1. Direct Personal Interviews:
 A face to face contact is made with the informants (persons from whom the information is
to be obtained) under this method of collecting data. The interviewer asks them questions
pertaining to the survey and collects the desired information. Thus, if a person wants to
collect data about the working conditions of the workers of the Tata Iron and Steel Company,
Jamshedpur, he would go to the factory, contact the workers and obtain the desired
information. The information collected in this manner is first hand and also original in
character.
2. Indirect Oral Interviews:
Under this method of data collection, the investigator contacts third parties generally called
witnesses who are capable of supplying necessary information. This method is generally
adopted when the information to be obtained is of a complex nature and informants are not
inclined to respond if approached directly. For example, when the researcher is trying to
obtain data on drug addiction or the habit of taking liquor, there is high probability that the
addicted person will not provide the desired data and hence will disturb the whole research
process. In this situation taking the help of such persons or agencies or the neighbours who
know them well becomes necessary. Since these people know the person well, they can
provide the desired data.

Though this method is very popular, its correctness depends upon a number of factors such as
1. The person or persons or agency whose help is solicited must be of proven integrity;
otherwise any bias or prejudice on their part will not bring out the correct information and the
whole process of research will become useless.
2. The ability of the interviewers to draw information from witnesses by means of appropriate
questions and cross-examination.
3. It might happen that because of bribery, nepotism or certain other reasons those who are
collecting the information give it such a twist that correct conclusions are not arrived at.

3. Information from Correspondents:

The investigator appoints local agents or correspondents in different places to collect

information under this method. These correspondents collect and transmit the information
to the central office where data are processed. This method is generally adopted by news
paper agencies. Correspondents who are posted at different places supply information
relating to such events as accidents, riots, strikes, etc., to the head office. The
correspondents are generally paid staff or sometimes they may be honorary correspondents
also. This method is also adopted generally by the government departments in such cases
where regular information is to be collected from a wide area. The biggest advantage of this
method is that, it is cheap and appropriate for extensive investigation. But a word of caution
is that it may not always ensure accurate results because of the personal prejudice and bias
of the correspondents. As stated earlier, this method is suitable and adopted in those cases
where the information is to be obtained at regular intervals from a wide area.

4. Mailed Questionnaire Method:

Under this method, a list of questions pertaining to the survey which is known as
Questionnaire is prepared and sent to the various informants by post. Sometimes the
researcher himself too contacts the respondents and gets the responses related to various
questions in the questionnaire. The questionnaire contains questions and provides space for
answers. A request is made to the informants through a covering letter to fill up the
questionnaire and send it back within a specified time. The questionnaire studies can be
classified on the basis of:
i. The degree to which the questionnaire is formalized or structured.
ii. The disguise or lack of disguise of the questionnaire and
iii. The communication method used.

When questionnaires are constructed in such a way that the objective is clear to the
respondents then these questionnaires are known as non- disguised; on the other hand,
when the objective is not clear, the questionnaire is a disguised one. On the basis of these
two classifications, four types of studies can be distinguished:
1. Non-disguised structured,
2. Non-disguised non-structured,
3. Disguised structured and
4. Disguised non-structured.

5. Schedules Sent Through Enumerators:

Another method of data collection is sending schedules through the enumerators or

interviewers. The enumerators contact the informants, get replies to the questions
contained in a schedule and fill them in their own handwriting in the questionnaire form.
There is difference between questionnaire and schedule. Questionnaire refers to a device for
securing answers to questions by using a form which the respondent fills in him self,
whereas schedule is the name usually applied to a set of questions which are asked in a
face-to face situation with another person. This method is free from most of the limitations
of the mailed questionnaire method.

Secondary Data:

The various sources of secondary data can be divided into two broad categories:
1. Published sources, and
2. Unpublished sources.

1. Published Sources:
The governmental, international and local agencies publish statistical data, and chief among
them are explained below:
(a) International Bublications:
There are some international institutions and bodies like I.M.F, I.B.R.D, I.C.A.F.E and U.N.O
who publish regular and occasional reports on economic and statistical matters.
 (b) Official Publications of Central and State Governments:
Several departments of the Central and State Governments regularly publish reports on a
number of subjects. They gather additional information. Some of the important publications
are: The Reserve Bank of India Bulletin, Census of India, Statistical Abstracts of States,
Agricultural Statistics of India, Indian Trade Journal, etc.
(c) Semi-Official Publications:
Semi-Government institutions like Municipal Corporations, District Boards, Panchayats, etc.
Publish reports relating to different matters of public concern.
(d) Publications of Research Institutions:
Indian Statistical Institute (I.S.I), Indian Council of Agricultural Research (I.C.A.R), Indian
Agricultural Statistics Research Institute (I.A.S.R.I), etc. Publish the findings of their research
programmes.
(e) Publications of various Commercial and Financial Institutions

(f) Reports of various Committees and Commissions appointed by the Government as the
Raj Committees Report on Agricultural Taxation, Wanchoo Committees Report on Taxation
and Black Money, etc. Are also important sources of secondary data.
(g) Journals and News Papers:
Journals and News Papers are very important and powerful source of secondary data.
Current and important materials on statistics and socio-economic problems can be obtained
from journals and newspapers like Economic Times, Commerce, Capital, Indian Finance,
Monthly Statistics of trade etc.

2. Unpublished Sources:
Unpublished data can be obtained from many unpublished sources like records maintained
by various government and private offices, the theses of the numerous research scholars in
the universities or institutions etc.

The Suitability Of Data:

The investigator must satisfy himself that the data available are suitable for the purpose of
enquiry. It can be judged by the nature and scope of the present enquiry with the original
enquiry. For example, if the object of the present enquiry is to study the trend in retail
prices, and if the data provide only wholesale prices, such data are unsuitable.

(A) Adequacy Of Data:

If the data are suitable for the purpose of investigation then we must consider whether the
data are useful or adequate for the present analysis. It can be studied by the geographical
area covered by the original enquiry. The time for which data are available is very important
element. In the above example, if our object is to study the retail price trend of india, and if
the available data cover only the retail price trend in the state of bihar, then it would not
serve the purpose.
(b) Reliability Of Data:
The reliability of data is must. Without which there is no meaning in research. The reliability
of data can be tested by finding out the agency that collected such data. If the agency has
used proper methods in collection of data, statistics may be relied upon.

It is not enough to have baskets of data in hand. In fact, data in a raw form are nothing but a
handful of raw material waiting for proper processing so that they can become useful. Once
data have been obtained from primary or secondary source, the next step in a statistical
investigation is to edit the data.Editing data collected from internal records and published
sources is relatively simple but the data collected from a survey need excessive editing.

While editing primary data, the following considerations should be borne in mind:

1. The data should be complete in every respect

2. The data should be accurate
3. The data should be consistent, and
4. The data should be homogeneous.
Data to posses the above mentioned characteristics have to undergo the same type of
editing which is discussed below:

5. Editing for Completeness:

while editing, the editor should see that each schedule and questionnaire is complete in all
respects. He should see to it that the answers to each and every question have been
furnished. If some questions are not answered and if they are of vital importance, the
informants should be contacted again either personally or through correspondence. Even
after all the efforts it may happen that a few questions remain unanswered. In such
questions, the editor should mark No answer in the space provided for answers and if the
questions are of vital importance then the schedule or questionnaire should be dropped.

(a) Editing for Consistency:

At the time of editing the data for consistency, the editor should see that the answers to
questions are not contradictory in nature. If they are mutually contradictory answers, he
should try to obtain the correct answers either by referring back the questionnaire or by
contacting, wherever possible, the informant in person. For example, if amongst others, two
questions in questionnaire are (a) Are you a student? (b) Which class do you study and the
reply to the first question is no and to the latter tenth then there is contradiction and it
should be clarified.
(b) Editing for Accuracy:
The reliability of conclusions depends basically on the correctness of information. If the
information supplied is wrong, conclusions can never be valid. It is, therefore, necessary for
the editor to see that the information is accurate in all respects. If the inaccuracy is due to
arithmetical errors, it can be easily detected and corrected. But if the cause of inaccuracy is
faulty information supplied, it may be difficult to verify it and an example of this kind is
information relating to income, age etc.
(c) Editing For Homogeneity:
Homogeneity means the condition in which all the questions have been understood in the
same sense. The editor must check all the questions for uniform interpretation. For example,
as to the question of income, if some informants have given monthly income, others annual
income and still others weekly income or even daily income, no comparison can be made.
Therefore, it becomes an essential duty of the editor to check up that the information
supplied by the various people is homogeneous and uniform.

Choice Between Primary and Secondary Data:

As we have already seen, there are a lot of differences in the methods of collecting Primary
and Secondary data. Primary data which is to be collected originally involves an entire
scheme of plan starting with the definitions of various terms used, units to be employed,
type of enquiry to be conducted, extent of accuracy aimed at etc. For the collection of
secondary data, a mere compilation of the existing data would be sufficient. A proper choice
between the type of data needed for any particular statistical investigation is to be made
after taking into consideration the nature, objective and scope of the enquiry; the time and
the finances at the disposal of the agency; the degree of precision aimed at and the status of
the agency (whether government- state or central-or private institution of an individual).

Now-a-days in a large number of statistical enquiries, secondary data are generally used
because fairly reliable published data on a large number of diverse fields are now available
in the publications of governments, private organizations and research institutions, agencies,
periodicals and magazines etc.

Experiments
Procedures Adopted In Experiments
Meaning Of Experiments
Research Design In Case Of Hypothesis Testing Research Studies
Basic Principles In Experimental Designs
Prominent Experimental Designs

The meaning of experiment lies in the process of examining the truth of a statistical
hypothesis related to some research problem. For example, a researcher can conduct an
experiment to examine the newly developed medicine. Experiment is of two types: absolute
experiment and comparative experiment. When a researcher wants to determine the impact
of a fertilizer on the yield of a crop it is a case of absolute experiment. On the other hand, if
he wants to determine the impact of one fertilizer as compared to the impact of some other
fertilizer, the experiment will then be called as a comparative experiment.

Research design can be of three types:

1. Research design in the case of descriptive and diagnostic research studies,

2. Research design in the case of exploratory research studies, and
3. Research design in the case of hypothesis testing research studies.
Here we are mainly concerned with the third one which is Research design in the case of
hypothesis testing research studies.

Research design in the case of hypothesis testing research studies:

Hypothesis testing research studies are generally known as experimental studies. This is a study
where a researcher tests the hypothesis of causal relationships between variables. This type of
study requires some procedures which will not only reduce bias and increase reliability, but will
also permit drawing inferences about causality. Most of the times, experiments meet these
requirements. Prof. Fisher is considered as the pioneer of this type of studies (experimental
studies). He did pioneering work when he was working at Rothamsted Experimental Station in
England which was a centre for Agricultural Research. While working there, Prof. Fisher found
that by dividing plots into different blocks and then by conducting experiments in each of these
blocks whatever information that were collected and inferences drawn from them happened to
be more reliable.

Nowadays, the experimental design is used in researches relating to almost every discipline of
knowledge. Prof. Fisher laid three principles of experimental designs:
1. The Principle of Replication
2. The Principle of Randomization and
3. The Principle of Local Cont
1. The Principle Of Replication:

According to this principle, the experiment should be repeated more than once. Thus, each
treatment is applied in many experimental units instead of one. This way the statistical
accuracy of the experiments is increased. For example, suppose we are going to examine the
effect of two varieties of wheat. Accordingly, we divide the field into two parts and grow one
variety in one part and the other variety in the other. Then we compare the yield of the two
parts and draw conclusion on that basis. But if we are to apply the principle of replication to
this experiment, then we first divide the field into several parts, grow one variety in half of
these parts and the other variety in the remaining parts. The entire experiment can be
repeated several times for better results.

2. The Principle of Randomization:

When we conduct an experiment, the principle of randomization provides us a protection
against the effects of extraneous factors. This means that this principle indicates that the
researcher should design or plan the experiment in such a way that the variations caused by
extraneous factors can all be combined under the general heading of chance. For example,
when a researcher grows one variety of wheat , say , in the first half of the parts of a field
and the other variety he grows in the other half, then it is just possible that the soil fertility
may be different in the first half in comparison to the other half. If this is so the researchers
result is not realistic. In this situation, he may assign the variety of wheat to be grown in
different parts of the field on the basis of some random sampling technique.
3. The Principle Of Local Control:
This is another important principle of experimental designs. Under this principle, the
extraneous factor which is the known source of variability is made to vary deliberately over
as wide a range as necessary. This needs to be done in such a way that the variability it
causes can be measured and hence eliminated from the experimental error. The experiment
should be planned in such a way that the researcher can perform a two-way analysis of
variance, in which the total variability of the data is divided into three components
attributed to treatments (varieties of wheat in this case), the extraneous factor (soil fertility
in this case) and experimental error.

2. Formal Experimental Design

(i) Completely randomized design:

This design involves only two principles i.e., the principle of replication and the principle of
randomization of experimental designs. Among all other designs this is the simpler and
easier because its procedure and analysis are simple. The important characteristic of this
design is that the subjects are randomly assigned to experimental treatments. For example,
if the researcher has 20 subjects and if he wishes to test 10 under treatment A and 10 under
treatment B, the randomization process gives every possible group of 10 subjects selected
from a set of 20 an equal opportunity of being assigned to treatment A and treatment B.
One way analysis of variance (one way ANOVA) is used to analyze such a design.

(ii) Randomized block design:

R. B. Design is an improvement over the C.R. design. In the R .B. Design, the principle of local
control can be applied along with the other two principles of experimental designs. In the
R.B. design, subjects are first divided into groups, known as blocks, such that within each
group the subjects are relatively homogenous in respect to some selected variable. The
number of subjects in a given block would be randomly assigned to each treatment. Blocks
are the levels at which we hold the extraneous factor fixed, so that its contribution to the
total variability of data can be measured. The main feature of the R.B. design is that, in this,
each treatment appears the same number of times in each block. This design is analyzed by
the two-way analysis of variance (two-way ANOVA) technique.

(iii) Latin squares design:

The Latin squares design (L.S design) is an experimental design which is very frequently used
in agricultural research. Since agriculture depends upon nature to a large extent, the
condition of research and investigation in agriculture is different than the other studies. For
example, an experiment has to be made through which the effects of fertilizers on the yield
of a certain crop, say wheat, are to be judged. In this situation, the varying fertility of the soil
in different blocks in which the experiment has to be performed must be taken into
consideration; otherwise the results obtained may not be very dependable because the
 output happens to be the effects of not only of fertilizers, but also of the effect of fertility of
soil. Similarly there may be the impact of varying seeds of the yield. In order to overcome
such difficulties, the L.S. design is used when there are two major extraneous factors such as
the varying soil fertility and varying seeds. The Latin square design is such that each fertilizer
will appear five times but will be used only once in each row and in each column of the
design. In other words, in this design, the treatment is so allocated among the plots that no
treatment occurs more than once in any one row or any one column. This experiment can be
shown with the help of the following diagram:

From the above diagram, it is clear that in L.S. design the field is divided into as
many blocks as there are varieties of fertilizers. Then, each block is again divided into as many parts as
there are varieties of fertilizers in such a way that each of the fertilizer variety is used in each of the block
only once. The analysis of L.S. design is very similar to the two-way ANOVA technique.

(iv) Factorial design:

Factorial designs are used in experiments where the effects of varying more than one factor are to be
determined. These designs are used more in economic and social matters where usually a large number
of factors affect a particular problem. Factorial designs are usually of two types:

(i) Simple factorial designs and

(ii) Complex factorial designs.

 Observation
Steps in Observation

Meaning and Characteristics of Observation

Types of Observation

Stages of Observation

Problems, Merits And Demerits

Observation is a method that employs vision as its main means of data collection. It implies the use
of eyes rather than of ears and the voice. It is accurate watching and noting of phenomena as they
occur with regard to the cause and effect or mutual relations. It is watching other persons behavior
as it actually happens without controlling it. For example, watching bonded labourers life, or
treatment of widows and their drudgery at home, provide graphic description of their social life and
sufferings. Observation is also defined as a planned methodical watching that involves constraints
to improve accuracy.

CHARACTERISTICS OF OBSERVATION

Scientific observation differs from other methods of data collection specifically in

four ways: (i) observation is always direct while other methods could be direct or
indirect; (ii) field observation takes place in a natural setting; (iii) observation
tends to be less structured; and (iv) it makes only the qualitative (and not the
quantitative) study which aims at discovering subjects experiences and how
subjects make sense of them (phenomenology) or how subjects understand
their life (interpretivism).

Lofland (1955) has said that this method is more appropriate for studying lifestyles or sub-cultures,
practices, episodes, encounters, relationships, groups, organizations, settlements and roles etc.
Black and Champion (1976) have given the following characteristics of observation:
Behavior is observed in natural surroundings.

It enables understanding significant events affecting social relations of the participants.

It determines reality from the perspective of observed person himself.

It identifies regularities and recurrences in social life by comparing data in our study with that of other
studies.

Statistical Analysis

1. Probability
2. Probability Distribution
2.1 Binomial Distribution
2.2 Poisson Distribution
2.3 Normal Distribution
3. Testing of Hypothesis
3.1 Small Sample
3.2 Large Sample Test
4. 2 Test

1. PROBABILITY
If an experiment is repeated under essentially homogeneous and similar conditions, two possible
conclusions can be arrived. They are: the results are unique and the outcome can be predictable and
result is not unique but may be one of the several possible outcomes. In this context, it is better to
understand various terms pertaining to probability before examining the probability theory. The main
terms are explained as follows:

(i) Random Experiment

An experiment which can be repeated under the same conditions and the
outcome cannot be predicted under any circumstances is known 94
as random experiment. For example: An unbiased coin is tossed. Here we are not in a position to
predict whether head or tail is going to occur. Hence, this type of experiment is known as random
experiment.

(ii) Sample Space

A set of possible outcomes of a random experiment is known as sample space. For example, in the
case of tossing of an unbiased coin twice, the possible outcomes are HH, HT, TH and TT. This can be
represented in a sample space as S= (HH, HT, TH, TT).
(iii) An Event
Any possible outcomes of an experiment are known as an event. In the case of tossing of an
unbiased coin twice, HH is an event. An event can be classified into two. They are: (a) Simple events,
and (ii) Compound events. Simple event is an event which has only one sample point in the sample
space. Compound event is an event which has more than one sample point in the sample space. In
the case of tossing of an unbiased coin twice HH is a simple event and TH and TT are the compound
events.

(iv) Complementary Event

A and A are the complementary event if A consists of all those sample point which is not included
in A. For instance, an unbiased dice is thrown once. The probability of an odd number turns up are
complementary to an even number turns up. Here, it is worth mentioning that the probability of
sample space is always is equal to one. Hence, the P (A) = 1 - P (A).

(v) Mutually Exclusive Events

A and B are the two mutually exclusive events if the occurrence of A precludes the occurrence of B.
For example, in the case of tossing of an unbiased coin once, the occurrence of head precludes the
occurrence of tail. Hence, head and tail are the mutually exclusive event in the case of tossing of an
unbiased coin once. If A and B are mutually exclusive events, then the probability of occurrence of A
or B is equal to sum of their individual probabilities. Symbolically, it can be presented as:

P (A U B) = P (A) + P (B)
If A and B is joint sets, then the addition theorem of probability can be stated as:
P (A U B ) = P(A) + P(B) - P(AB)

Addition Theorem of Probability

Let A and B be the two mutually exclusive events, then the probability of A or B is equal to the sum of
their individual probabilities. (for detail refer mutually exclusive events)

Multiplication Theorem of Probability

Let A and B be the two independent events, then the probability of A and B is equal to the product of
their individual probabilities. (for details refer independent events)

2. PROBABILITY DISTRIBUTION
If X is discrete random variable which takes the values of x1, x2,x3.. Xn and the corresponding
probabilities are p1, p2, .pn, then, X follows the probability distribution. The two main properties of
probability distribution are: (i) P(Xi) is always greater than or equal to zero and less than or equal to one,
and (ii) the summation of probability distribution is always equal to one. For example, tossing of an
unbiased coin twice.

Then the probability distribution is:

X (probability of obtaining head): 0 12
P(xi) :
Expectation of probability
Let X be the discrete random variable which takes the value of x 1, x2, xn then the respective
probability is p1, p2, pn, then the expectation of probability distribution is p1x1 + p2x2 +
.. + pnxn. In the above example, the expectation of probability distribution is (0*
+1*1/2+2*) =1.

2.1 BINOMIAL DISTRIBUTION

The binomial distribution also known as Bernoulli Distribution is associated with

the name of a Swiss mathematician, James Bernoulli who is also known as Jacques
or Jakon (1654 1705). Binomial distribution is a probability distribution
expressing the probability of one set of dichotomous alternatives. It can be
explained as follows:
i. If an experiment is repeated under the same conditions for a fixed number of
trials, say, n.
ii. In each trial, there are only two possible outcomes of the experiment. Let us
define it as success or failure. Then the sample space of possible outcomes of
each experiment is:
iii. S = [failure, success]
iv. The probability of a success denoted by p remains constant from trial to trial
and the probability of a failure denoted by q which is equal to (1 p).
v. The trials are independent in nature i.e., the outcomes of any trial or sequence
of trials do not affect the outcomes of subsequent trials. Hence, the multiplication
theorem of probability can be applied for the occurrence of success and failure.
Thus, the probability of success or failure is p.q.
vi. Let us assume that we conduct an experiment in n times. Out of which x times
be the success and failure is (n-x) times. The occurrence of success or failure in
successive trials is mutually exclusive events. Hence, we can apply addition
theorem of probability.
vii. Based on the above two theorems, the probability of success or failure is
P(X) = nCxpxqn-x

n!
--------------- . Px qn-x
x ! (n x) !
where p = probability of success in a single trail, q = 1 p, n = Number of trials and
x = no. of successes in n trials.
2.2 POISSON DISTRIBUTION

Poisson distribution was derived in 1837 by a French Mathematician Simeon D

Poisson (1731 1840). In binomial distribution, the values of p and q and n are
given. There is a certainty of the total number of events. But there are cases
where p is very small and n is very large and such case is normally related to
poisson distribution. For example, persons killed in road accidents, the number of
defective articles produced by a quality machine. Poisson distribution may be
obtained as a limiting case of binomial probability distribution, under the
following condition.
i. P, successes, approach zero (p 0)
ii. np = m is finite.
The poisson distribution of the probabilities of occurrence of various rare events
(successes) 0,1,2,. Are given below:

2.3 NORMAL DISTRIBUTION