Talk:Coverage probability
 | This article is rated Stub-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
| |||||||||||
.svg){kind=small} | Daily pageviews of this article
A graph should have been displayed here but graphs are temporarily disabled. Until they are enabled again, visit the interactive graph at pageviews.wmcloud.org |
Credible Intervals edit
This article should also talk about the interpretation of Coverage probability of Credible Intervals that are generated by Bayesian methods MATThematical (talk) 03:40, 24 February 2010 (UTC)
Domain (sigma-algebra) of the coverage probability edit
Hi!
Specifying the domain of the coverage probability might improve the clarity of the article (I couldn't find it specified anywhere in the literature).
Kindest regards PodobnikT (talk) 12:16, 16 April 2012 (UTC)
- Hi! Specifying sigma-algebra, underlying the coverage probability, would, for example, help (at least me) to understand whether or not the coverage probability is truly a probability, and whether or not the confidence distribution is an example of the coverage probability. Kindest regards, PodobnikT (talk) 08:29, 17 April 2012 (UTC)
- It should be possible to do what you want with simpler mathematics than measure theory. I invite you to look at Robust statistics#Empirical influence function for an example of how non-understable something can be made by unnecessary sophistication. However, you may find that the article Random compact set treats things at the level you mention. But there is a simple mathematical definition at Confidence interval#Approximate confidence intervals that may serve. The article confidence distribution should/might have something relevant to you if you haven't already seen that. Melcombe (talk) 12:52, 17 April 2012 (UTC)
Possible error in definition of 'coverage probability'. edit
 | This article's factual accuracy is disputed. |
The following statement may be inaccurate: "The coverage probability is the actual probability that the interval contains the true mean remission duration in this example."
Once an interval is constructed, it either contains the population parameter (e.g. mean) or it does not. Hence, the actual probability of a constructed interval being correct is either 0 or 1. That seems inconsistent with the notion of nominal and actual probabilities being equal when assumptions are met.