What is the most suitable estimation for hypothesis testing?
For hypothesis testing, point estimation (that is , a statistical test that rejects or do not reject a hypothesis) is better than a confidence interval. A 95 % confidence interval gives you 95 % probability that the true estimate lies between two numbers. There is a relation between the two. You could use your confidence interval to reject or don’t reject your null hypothesis.