ABSTRACT
In the analysis of random data, we encounter situations to check various
statistical models or “hypotheses” against the data. The usual situation is to test
or check a particular probability distribution for consistency in representing data
from a certain experiment. The hypothesis that this distribution is the right one is
called null hypothesis, often denoted by H
: This hypothesis may have emerged
from long experience associated with an experiment and it is necessary to see
whether the hypothesis is still correct when there has been some changes in
circumstances which call it into question. Alternatively, the hypothesis may be the
result of a theoretical analysis or a logical argument and the theory needs to be
verified.
