A hypothesis can be tested. Any claim that can be tested will have a hypothesis about the outcome. A dog owner said his dog paid more attention to the morning paper if there was a cat in it. The owner wants to see if dogs will read the newspaper more often if there is a cat above the fold line. A null hypothesis is a statistical hypothesis test that assumes a certain outcome. The hypothesis is that the presence of a cat on the newspaper will not increase the likelihood of a dog reading it. An alternative hypothesis is the opposite of a null hypothesis and covers any effects the cat may have on the dog.
The idea of a null hypothesis is that an experiment is conducted to try to disprove it. To fully support a hypothesis, there needs to be a p-value that shows the likelihood that the result was due to the variables and not to chance. The p-value is called statistical significance. The p-value is the statistical power of a hypothesis test. The higher the statistical power, the less likely the error is. When the null hypothesis is true, a higher power shows the probability of accepting the alternative hypothesis. The problem arises when an experiment is conducted with low statistical power because a conclusion could be false.
In order for the result to be accepted, statistical power of 80 percent or more is needed. There is only a 20 percent probability of an error with an 80 percent power.
You forgot something?
You neglected something?