1
$\begingroup$

Could somebody explain to me the relationship between coverage error and type one errors in multiple comparisons testing, if there is one in fact? Does a coverage error occur when the true value of the comparison is in fact not within the parameters of the confidence interval? Further, does this have any influence on the probability of a type I/II error for that comparison?

Thanks

$\endgroup$
1
$\begingroup$

"Coverage error" at least as I have seen it, refers to whether the e.g. 95% CI is, in fact, a 95% CI or is too liberal or too conservative. If there is a coverage error, then the type I error rate is not as stated, but it could be either too low or too high, depending on the type of error.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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