Humberto Barreto wrote:
>
> Hi,
>
> The statement was:
> "there is a 95% chance that the true population parameter is within the
> computed interval."
(etc)
But that *wasn't* the statement, that was only *part* of the statement.
The original statement also described a sample being gathered; it was
not made clear whether the probability was conditional on the results of
the sample or not. If so, the statement was wrong; if not, correct.
The statement was lamentably vague (as discussions of conditional
probability, etc, in colloquial language tend to be). I would go with
"absolutely not well written" but not with "absolutely not true". The
latter requires an unambiguous interpretation [or at least that whatever
interpretation we pick the same result follows.]
-Robert Dawson
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
