In research problem hypothesis occupies an important place and position. It is the hypothesis which is to be accepted refuted proved or disproved. All the data is to be collected with an eye on the hypothesis.
image source: prconversations.com
It is, therefore, essential to know how to formulate hypothesis. A hypothesis can be developed in the initial stages but as the investigations proceed the original hypothesis may undergo numerous changes in some cases there may be even substantial changes.
ADVERTISEMENTS:
The reason for the same being not far to seek, as new accumulative evidences not anticipated at the commencement stage may come to light. In the words of Robertson and Wright, Even in the case of most carefully planned research project, some hypothesis are not likely to be formed until the evidence, obtained to test other hypothesis is being analysed. Sometimes hypothesis developed in the course of analysing evidence necessitates the collection of additional Data.”
It is always in the interest of researcher that he should keep his mind open and flexible. He should be prepared to retrace the steps and to revise his hypothesis, if need be. In some cases, after the collection of data, it may even become necessary to abandon the original hypothesis.
In some other cases hypothesis may take formulative and definitive stage only after some data has been collected.
Formulation of careful hypothesis becomes necessary because that makes the investigation easy. There can be number of hypothesis and some may be discarded in the very beginning while others in the course of investigation.
ADVERTISEMENTS:
Hypothesis should be such that it results in efforts to discover something not already known: That is why it is also called creative art. Important considerations for workable hypothesis have already been discussed and care should be taken to see that the hypothesis is or problem of study is not only interesting but also useful both for the society as well as the research.
Alternative Hypothesis:
A researcher has a problem before him which he wants to investigate and he starts with a hypothesis to solve that. His research findings may or may not prove that. A researcher should, therefore, start with trying to determine all the alternative means. He should try to determine which of the alternative courses of action or solutions or explanations can be applied to the problem. While finding out an alternative the researcher will, of courses, have to think in terms of money, predictability, manpower, area to be covered, etc.
Certain important steps are involved in the formation of alternative hypothesis. A measure of efficiency applicable to all the alternative courses of action is to be selected in the first instance. After this has been done a set of accepted conditions for each alternative course of action is to be assigned.
The conditions are then reformulated as hypothesis which are mutually exclusive and jointly exhaustive. If acceptance of one alternative hypothesis rather than another will make no difference whatsoever in subsequent behaviour than the problem or its formulation is meaningless.
ADVERTISEMENTS:
It will have to be accepted that in social research there is not scientific way of selecting one of the alternative hypothesis as valid unless there is an index of efficiency which can be applied to each of the alternative course of action. If there are two alternative hypotheses, there should be one point of agreement and one point of disagreement. All the alternative hypotheses should be exhaustive and mutually exclusive, failing which it shall not be possible to select one alternative.
Null Alternative:
For statistical testing of hypothesis both alternative and null hypothesis are involved. Null hypothesis in its simplest form means that there is no difference between two populations in respect of same property and that the difference, it any, is only incidental and insignificant. It means in null hypothesis it is presumed that both the methods being used in the study are good.
When one method is being preferred over the other it is known as null alternative. In other words, null alternative is a principle which states that a person is innocent unless he is proved guilty. A null hypothesis asserts that results found in research do not differ significantly from the expected results on a probability basis. A null hypothesis is exact and useful because it is easier to disprove the contrary of a hypothesis than to prove it, with complete certainty.
Both null hypothesis and alternative hypothesis are chosen before the sample is drawn. While null hypothesis is being chosen it means that the researcher is interested in disproving something. Such a hypothesis should not state approximately of a certain value.