In: Statistics and Probability
p-values are used to reject or accept the null hypothesis based on 2 factors:
1. Confidence interval we have choosen
2. Whether it is a two tailed or single tail test
Example to demonstrate the same:
Confidence interval is 90%
Ho : There is no strong relationship between A and B
H1: There is strong relationship between A and B
Clearly the above Hypothesis test is based on two tailed test.
So the null hypothesis will be rejected only if the calculated test statistic p-value is greater than 5% (since it is a two tailed test).
This is the basic understanding of accepting/rejecting null hypothesis.