Subject: Re: Unfair TV Ratings - Ron Reagan show (TP/30Something)
From: unknown@Apple.COM (Matt Ackeret)
Date: 1991-08-28, 14:13
Newsgroups: rec.arts.tv,alt.tv.twin-peaks

In article <1991Aug28.054403.7162@risky.ecs.umass.edu> giovin@risky.ecs.umass.edu (Rocky J Giovinazzo) writes:
> >Did anyone see the recent Ron Reagan talk show featuring 
> >Bob Iger of ABC, Steven Bocchco, Diane English and some other
> >tv people?  Apparently, there are only 4000 Nielson households in the
> >country (USA)!  I thought it was 1/1000 households but its only 1/23000 !!!
> >Holy Cow!  (The guests admitted that this representation was extremely
> >low but didn't seem to really care much -- except maybe Diane English,
> >creator of Murphy Brown).  

I didn't see the show, but I do know that it can be shown 
mathematically that a seemingly small sample if chosen scientifically (i.e.
represents the whole well) can very accurately represent the whole group.

That is, a tiny proportion can give results closely approximating
the results that would be gotten if everyone was sampled.

I'm not saying that the Nielsen families ARE a statistically valid
sample.. they probably aren't. But simply saying it's 1/23000-th of the 
population doesn't necessarily mean it's going to give incorrect results.

Someone else can probably give out a more thorough explanation.
-- /I'm just a summer contractor. Don't take any of my personal ramblings as \ |official Apple ramblings. unknown@cats.ucsc.edu Apple IIGS Forever!| |Matt_Ackeret@gateway.qm.apple.com unknown@ucscb.ucsc.edu unknown@apple.com | \To get ULTIMA VI GS or ARMOR ALLEY GS written-mail me. CHEAP CD info-mail me./