## What is Hypothesis?

Hypothesis is a proposition which can be put to a test to determine validity and is useful for further research.

Hypothesis is a statement which can be proved or disproved. It is a statement capable of being tested. In a sense, hypothesis is a question which definitely has an answer. Hypothesis aids us a great deal while collecting, tabulating and analyzing data and other relevant information.

Table of Contents

- 1 What is Hypothesis?
- 2 Hypothesis Definition
- 3 Meaning of Hypothesis
- 4 Characteristics of Hypothesis
- 4.1 Conceptual Clarity
- 4.2 Need of the empirical referents
- 4.3 Hypothesis should be specific
- 4.4 Hypothesis should be within the ambit of the available research techniques
- 4.5 Hypothesis should be consistent with the theory
- 4.6 Hypothesis should be concerned with observable facts and empirical events
- 4.7 Hypothesis should be simple

- 5 Formulation of Hypothesis
- 6 Null Hypothesis
- 7 Test of Hypothesis

Hypothesis thus is inevitable in any kind of research, if it is to be carried out successfully. The meaning and exact nature of hypothesis will become clear from the following definitions.

## Hypothesis Definition

## Meaning of Hypothesis

From the above mentioned definitions of hypothesis, its meaning can be explained in the following ways:

- At the primary level, a hypothesis is the possible and probable explanation of the sequence of happenings or data.
- Sometimes, hypothesis may emerge from an imagination, common sense or a sudden event.
- Hypothesis can be a probable answer to the research problem undertaken for study.
- Hypothesis may not always be true. It can get disproven. In other words, hypothesis need not always be a true proposition.
- Hypothesis, in a sense, is an attempt to present the interrelations that exist in the available data or information.
- Hypothesis is not an individual opinion or community thought. Instead, it is a philosophical means which is to be used for research purpose. Hypothesis is not to be considered as the ultimate objective; rather it is to be taken as the means of explaining scientifically the prevailing situation.

## Characteristics of Hypothesis

Not all the hypotheses are good and useful from the point of view of research. It is only the few hypotheses satisfying certain criteria that are good, useful and directive in the research work undertaken.

The characteristics of a good hypothesis can be listed as below.

- Conceptual Clarity
- Need of the empirical referents
- Hypothesis should be specific
- Hypothesis should be within the ambit of the available research techniques
- Hypothesis should be consistent with the theory
- Hypothesis should be concerned with observable facts and empirical events
- Hypothesis should be simple

### Conceptual Clarity

The concepts used while framing hypothesis should be crystal clear and unambiguous. Such concepts must be clearly defined so that they become lucid and acceptable to everyone.

How are the newly developed concepts interrelated and how are they linked with the old one is to be very clear so that the hypothesis framed on their basis also carries the same clarity. A hypothesis embodying unclear and ambiguous concepts can to a great extent undermine the successful completion of the research work.

### Need of the empirical referents

A hypothesis can be useful in the research work undertaken only when it has links with some empirical referents. Hypothesis based on moral values and ideals are useless as they cannot be tested. Similarly, hypothesis containing opinions as good and bad or expectation with respect to something are not testable and therefore useless.

For example, ‘current account deficit can be lowered if people change their attitude towards gold’ is a hypothesis encompassing expectation. In case of such a hypothesis, the attitude towards gold is something which cannot clearly be described and therefore a hypothesis which embodies such an unclear thing cannot be tested and proved or disproved. In short, the hypothesis should be linked with some testable referents.

### Hypothesis should be specific

For the successful conduction of research, it is necessary that the hypothesis is specific and presented in a precise manner. Hypothesis which is general, too ambitious and grandiose in scope is not to be made as such hypothesis cannot be easily put to test. A hypothesis is to be based on such concepts which are precise and empirical in nature. A hypothesis should give a clear idea about the indicators which are to be used.

For example, a hypothesis that economic power is increasingly getting concentrated in few hands in India should enable us to define the concept of economic power. It should be explicated in terms of the measurable indicator like income, wealth, etc. Such specificity in the formulation of hypothesis ensures that the research is practicable and significant.

### Hypothesis should be within the ambit of the available research techniques

While framing the hypothesis, the researcher should be aware of the available research techniques and should see that the hypothesis framed is testable on the basis of them.

In other words, a hypothesis should be researchable and for this, it is important that a due thought has been given to the methods and techniques which can be used to measure the concepts and variables embodied in the hypothesis.

It does not, however, mean that hypotheses which are not testable with the available techniques of research are not to be made. If the problem is too significant and therefore the hypothesis framed becomes too ambitious and complex, it’s testing becomes possible with the development of new research techniques or the hypothesis itself leads to the development of new research techniques.

### Hypothesis should be consistent with the theory

A hypothesis must be related to the existing theory or should have a theoretical orientation. The growth of the knowledge takes place in the sequence of facts, hypothesis, theory and law or principles.

It means the hypothesis should have a correspondence with the existing facts and theory. If the hypothesis is related to some theory, the research work will enable us to support, modify or refute the existing theory. Theoretical orientation of the hypothesis ensures that it becomes scientifically useful.

According to Prof. Goode and Prof. Hatt, research work can contribute to the existing knowledge only when the hypothesis is related to some theory.

### Hypothesis should be concerned with observable facts and empirical events

This enables us to explain the observed facts and situations and also verify the framed hypothesis.

In the words of Prof. Cohen and Prof. Nagel, “hypothesis must be formulated in such a manner that deduction can be made from it and that consequently a decision can be reached as to whether it does or does not explain the facts considered.

### Hypothesis should be simple

If the research work based on a hypothesis is to be successful, it is necessary that the later is as simple and easy as possible. An ambition of finding out something new may lead the researcher to frame an unrealistic and unclear hypothesis. Such a temptation is to be avoided.

Framing a simple, easy and testable hypothesis requires that the researcher is well acquainted with the related concepts.

## Formulation of Hypothesis

The real beginning of any research is made with the formulation of hypothesis. In a sense, research is nothing but accepting the hypothesis by proving it or rejecting it if it is disproved or modifying it.

Moreover, in any type of research work, the information and data is to be collected with reference to the hypothesis and the concepts embodied in it. Hypothesis therefore occupies an important place in any type of research.

Formulation of hypothesis, however, requires that the difficulties encountered are overcome. A researcher may suffer from a number of difficulties at the stage of formulating a good hypothesis

- The researcher should have a thorough knowledge of the accepted theories and basic concepts of that research area where he has decided to work in.
- The researcher should also acquire the logical and scientific thinking power to frame a hypothesis based on the theories and basic concepts known to him.
- The researcher should also be well acquainted with the available research methods and techniques.

Normally, the hypothesis made in the beginning of research is of crude or working nature. Such a working hypothesis is to be made while planning a research work. As the research work proceeds with the working hypothesis, new information, data and evidence becomes available. In the light of new information and evidence, the working hypothesis is to be modified and revised.

Sometimes, the working hypothesis changes in a significant way after the modifications are made. In some researches, the hypothesis is formulated not in the beginning but at the time of classification and analysis of data and information.

In the case of such a hypothesis also it becomes necessary that new or additional information is collected. It thus implies that every hypothesis is subject to change. In order to put the research work in an operative mode, several alternative hypotheses are made in the beginning.

While framing such hypotheses utmost care is to be taken while using the concepts. The nature of the hypothesis should be such that it enables the researcher to find out something new, something which is previously unknown.

In the context of research work and while performing the hypothesis testing exercise, both the alternative hypothesis which is to be proved and accepted and null hypothesis, which is to be disproved, are important and required.

The main hypothesis of the research work is the research hypothesis or the alternative hypothesis. Researcher’s job is to collect information and data so as to prove the alternative hypothesis so that it can be accepted. Null hypothesis on the other hand is the exact opposite of research or alternative hypothesis.

Null hypothesis is also called a hypothesis with no difference. Like the research or alternative hypothesis, the null hypothesis is also a statement.

The logic behind formulating a null hypothesis is that it is always easy to prove that a statement is wrong than to prove that a statement (research hypothesis) is cent percent true.

In short, while framing hypothesis for research work, it is important that at least two hypotheses are framed, one of which is a null hypothesis and the other one is an alternative hypothesis.

For instance, a null hypothesis and alternative hypothesis can be as below.

## Null Hypothesis

The average age of entry in to the labour market of commerce graduates is 22 years.

However, the collected data and information, when analysed, reveals that Hypothesis the average entry age is greater than or less than 22 years, then the null hypothesis gets rejected.

**In such a case the alternative hypothesis can be as under**

- The average age of entry into the labour market is greater than 22 years (> 22)
- The average age of entry into the labour market is less than 22 years (< 22)
- The average age of entry into the labour market is not 22 years (‘“ 22)

## Test of Hypothesis

As stated in the beginning, the hypothesis formulation marks the beginning of any research. After the hypothesis is formulated in the context of a research problem, next process involves a collection of relevant data and information and analysis of the same using an appropriate statistical technique, which proves or disproves the hypothesis formulated in the beginning.

The testing of hypothesis thus represents the end of the research work. Testing of hypothesis can be considered as the most important step in any type of research work as it determines the fruitfulness of the research work.

Unless the hypothesis is tested, it will only remain an inference or a proposition. The act of determining the validity of the hypothesis based on the collected data is called the testing of hypothesis.

The exercise of hypothesis testing is a systematic work and normally involves following stages or steps:

- Stating the hypothesis of interest
- Collection of relevant data and information
- Formation of null hypothesis
- Alternative Hypothesis
- Selection of suitable test statistic
- Determine the level of significance
- Decision

### Stating the hypothesis of interest

Based on the research problem and a primitive understanding of the relationship between the variables involved, a researcher formulates a hypothesis of interest or a research hypothesis which he wants to prove.

### Collection of relevant data and information

Given the research problem and the formulated hypothesis of interest, the next step is to collect the relevant data and information to proceed further towards the end objective (i.e. proving the research hypothesis).

### Formation of null hypothesis

For the testing purpose, a null hypothesis is formed based on the statistical data. The null hypothesis is also called as the hypothesis with no difference.

In other words, null hypothesis states that there is no difference between the variables involved in the hypothesis or the variables are not related.

For example, if the research hypothesis is that the commerce graduates are more employable than the arts graduates, then the null hypothesis will be that both are equally employable or that there is no difference in the employment opportunities available to both.

If in research hypothesis, price and demand are said to be inversely related, the null hypothesis assumes them independent or states that price and demand are not related.

### Alternative Hypothesis

After the formulation of null hypothesis, alternative hypothesis can be derived. Alternative hypothesis is the negation of null hypothesis and can be more than one and conform to the research hypothesis.

In the example of employability, the alternative hypothesis can be

**Two-sided**- commerce graduates are more employable or arts graduates are more employable

**One-sided**- commerce graduates are having more employability
- arts graduates are having more employability.

### Selection of suitable test statistic

The next step in the hypothesis testing exercise is that of selecting an appropriate statistical test. It can be chi-square test, t-test or f-test or any other test. Such a test is carried out at a given level of significance.

### Determine the level of significance

As stated in the above step a statistical test is conducted at a given level of significance

- A level of significance indicates the probability of rejecting or accepting the null hypothesis.

### Decision

The last step in testing hypothesis is that of taking a decision on the basis of the given level of significance

- It is seen whether the null hypothesis falls in the accepting region or in rejecting region and accordingly a decision is taken. In this way, the acceptance or rejection of null hypothesis determines the acceptance or rejection of the initial research hypothesis.