Why is it necessary to have a significance threshold when performing the hypothesis tests?
Whenever we have to check hypothesis test, we can never claim that the conclusion we are drawing is 100% true, since, there always will be an error.
So, we pre-decide a significance threshold level up to which we can bear the error.
For example, you are claiming with 5% significance threshold that the mean age of students based on a sample of students of your class is 18 ( say )
That mean, you are 95% confidence that the mean age of all the students of your class will be 18
and 5% is the amount of error, so the probability of mean age being more than 18 or less than 18 is 0.05
Why pre decided?
If this is not pre decided some times we might come up with a solution which is not the true representative of the population
significance level is basically type I error probability = P[ Rejecting null hypothesis | it is true ]
Get Answers For Free
Most questions answered within 1 hours.