1
$\begingroup$

When the point estimator is relatively small numbers, a confidence interval is likely to contain zero, be fairly wide and include both positive and negative values. Basically, when CI contains zero we would say that we cannot reject null hypothesis, but is there another conclusions that we can get from CI? In that case what additional information (interpretation) can provide us the confidence interval?

$\endgroup$

marked as duplicate by whuber Jun 14 '16 at 20:36

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • $\begingroup$ Search our site for confidence interval to obtain a full answer. $\endgroup$ – whuber Jun 14 '16 at 20:36
  • $\begingroup$ It rather depends on your hypothesis. if you were testing a product for safety, you would want to ensure a 0% fatality rate (or to put it another way that it was 100% safe). In this case you would be happy with a confidence interval that includes zero and would ignore the negative side of the confidence interval. $\endgroup$ – Robert Jones Jun 15 '16 at 18:03
1
$\begingroup$

The main purpose is to inform you about out confidence in the value of the estimator. If confidence interval is very wide, it means that we have little confidence in the value. If the value is small, the we can't even say whether it's zero or not. That's it.

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.