Bring out the conditions of a valid hypothesis.

A good/valid hypothesis should be a logically consistent and practically workable explanation of the given phenomenon. It should explain the facts that we analyze. Hence, a valid hypothesis should be consistently explanatory. A valid hypothesis should be verifiable with reference to empirical facts. A good hypothesis should be the basis of further deductive inferences. If a hypothesis states that A is the cause of B, it must be the ground for predicting B from any further instance of A. A hypothesis is considered good only if it satisfies the standards of relevance. Hence, a good hypothesis should be purposeful and useful in solving a problem. For example, the knowledge of insulin deficiency as the cause of diabetes has made possible the effective treatment of the disease. A vague hypothesis cannot be a good one and hence a good hypothesis must be clear and conceivable. Finally, a valid hypothesis is not supposed to goagainst previously established truths in the field of knowledge. Copi and Cohen have identified three conditions to decide the merit of rival hypotheses:

i) Compatibility with previously established hypotheses

ii) Predictive power and

iii) Simplicity