|
The Sea to Shining See Concert to benefit people hurt by the September
11 terrorist disaster was a TV special featuring a large number of
well-known musicians and actors. A prediction was made by Paul Bethke.
On Fri, 28 Sep 2001, Paul wrote:
> Did you give any prediction to the 2-hour "Telethon" that aired
> last Friday on *every* channel from "Sea to shining see"?
> (I like that title.) If not - I'd like to make one.
>
> I'm still not sure how you do those duration-predictions. I'd
> think the two hours of the show (8:00-10:00 Central) would be
> a start - but I would think something like this could linger
> after anyone watching got "worked up". What do you think?
I think that is a good idea, and it had not occurred to me.
My first reaction is to take the time you give for the program,
8-10 PM = 20:00-22:00, CDT; +5 -> 25-27 -> 01:00 to 03:00 UTC,
Saturday 22 Sept 2001. Positive cumdev of Chisquare, seconds
resolution, medium expectation. That could be the formal
prediction, with exploration of linger as an option.
> Ok. That sounds logical. I *do* like the "linger", however.
> Let's be sure to explore that option.
The following graph shows the formal result, which has Chisquare
of 7409.6 on 7200 degrees of freedom with p = 0.041.
The next graph explores the "linger" effect. Though it is not a robust
interpretation, the graph does show that the trend during the show
continued for about an hour after it ended.
|