On Mon, 17 Mar 2003 16:21:45 GMT, Syrahz Derzai <[EMAIL PROTECTED]> wrote:
>Kevin J <[EMAIL PROTECTED]> wrote: >> > "plasma C level was found to be between Y and Z at X level >> > of confidence" is not the information a confidence interval provides. > >> Umm, you are wrong. Ever notice how polling data is written? It's a >> way of writing a CI, a way to write an interval calculated from sample >> data the includes the parameter with a preassigned level of >> confidence. Meaning, in simple english that given say a 95% CI, there >> is a 95% chance the population parameter falls within the 95% CI >> calculated from the sample data. > >This is the heart of the problem. I sympathize with researchers who refuse >to calculate CIs, because 99% of the time it will be interpreted the same >way you do. CI does not imply that the population parameter is between Y and >Z with X probability. To make statements like this you would need to >consider a prior parameter distribution and combine it with the likelihood >of the data. OK, technically, the 95% confedence interval refers to our confidence in the procedure (the construction of the CI). My understanding of this was that 95 times out of 100, a 95% CI constructed from the sample data will enclose the population parameter. I realize there is a distinction between saying this and saying that there is a 95% chance that the population parameter will fall within a _particular_ CI, but I had always thought this distinction very slight. It appears I am wrong. Both of my stats texts do emphasize that there is a distinction, but don't explain what the real world impact of this is. Care to educate me? I apologize for my previous attitude - it's an acceptable way of posting in MFW, but I didn't see the cross-posts. Who cross-posted this to a stats group? :) -- Kevin J hi spambot, my e-mail adress was made especially for YOU! . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
