Basic Concept of Null Hypothesis
What is a Null Hypothesis?
A very simple understanding of a null hypothesis states that is contrary to the alternative hypothesis which is denoted by H1. However the main idea behind this hypothesis is more complicated than this. Basically this hypothesis is the one that the experimenter attempts to reject, nullify or disprove. The word null in this term sometimes links with the common opinion regarding something, whereas the alternate hypothesis is the one that the experimenter actually thinks is the reason behind a phenomenon. The conclusion of an experiment will always tie with this hypothesis by accepting or rejecting it instead of the alternate hypothesis. Irrespective of this, a number of experimenters ignore this hypothesis during the process of hypothesis testing that is a poor practice and can result in bad impacts.
Null Hypothesis and Significance Tests
If there is a significance tests which is able to produce a 99% or 95% probability that the outcomes are not able to fit in the null hypothesis then this hypothesis is rejected or nullified with respect to the alternate hypothesis. And other case, the null is accepted. These assumptions are the only ones that hold true and it is not accurate to accept or reject the alternative hypothesis. When a researcher accepts a null it does not implies that this hypothesis is true. Still it remains a hypothesis and falls under the rule of falsifiability in the similar manner as when the null is rejected and the alternative is not proved.
Issues With the Null Hypothesis
There are a number of issues associated with the null hypothesis. One of the main issues is that a number of experimenters and researchers observe acceptance of a null as an experiment that has failed. This actually is a poor observation as rejection or acceptance of any of the hypothesis is basically a positive outcome. Even if the case where the H0 is not refuted there is something new which is learned in this case by the world of science. This means that the failure term should only be used for the errors that occur in the design of the experiments or in the case of inappropriate assumptions made in the beginning of the experiment.