The scientific method is the process of gaining knowledge, improving previous theory, developing a new theory, all of which depends upon the accuracy of the results. The accuracy along with frequency and nature is important for leading to a hypothesis.
All of the above mentioned considerations have to be noticed for identification of a hypothesis. These steps constitute the data analysis and ultimately lead to inference.
The first step is to check whether the observations differ qualitatively (difference in properties, objects, charcteristics) or quantitatively (in numbers). Then, check the similarity of the observation with that of the previously available data. After that the degree of dissimilarity and the frequency of the same is cosidered.