VoyForums
[ Show ]
Support VoyForums
[ Shrink ]
VoyForums Announcement: Programming and providing support for this service has been a labor of love since 1997. We are one of the few services online who values our users' privacy, and have never sold your information. We have even fought hard to defend your privacy in legal cases; however, we've done it with almost no financial support -- paying out of pocket to continue providing the service. Due to the issues imposed on us by advertisers, we also stopped hosting most ads on the forums many years ago. We hope you appreciate our efforts.

Show your support by donating any amount. (Note: We are still technically a for-profit company, so your contribution is not tax-deductible.) PayPal Acct: Feedback:

Donate to VoyForums (PayPal):

[ Login ] [ Contact Forum Admin ] [ Main index ] [ Post a new message ] [ Search | Check update time | Archives: 1, 2, 3, [4] ]

[ Next Thread | Previous Thread | Next Message | Previous Message ]

Date Posted: 13:18:11 10/29/02 Tue
Author: tim
Subject: Why the Nielsen Corp. does it that way...(O/T)
In reply to: CW 's message, "Re: Further..." on 19:50:52 10/14/02 Mon

The Nielson is such a tiny sample of the population that it should be going the way of the dodo. Although the Tivo data covers only Tivo subscribers it does include every last one of them, a much larger group of people than the whole Nielson sample. From my understanding the Nielson boxes only record what the television set is actually displaying not what any separate recorder may be doing near the set. Since Firefly is in the top five of all shows recorded in Tivo data, the Nielson data is not giving a correct picture of how many people intend to see it.

The social scientist in me wants to scream and hit and throw things when I hear stuff like this. So I apologize if the following sounds preachy and/or didactic. It's not your fault that you stumbled onto a major pet peeve of mine: namely, your assertion that Nielsens "should go the way of the dodo" because of the small sample size.

Let me tell you a little story:

In 1936, as the presidential election approached, the magazine Literary Digest polled over two million voters and predicted a huge win for Republican Alf Landon. Given that the vast majority of Americans think Alf is just a puppet who hawks phone services with other "celebrities," I think you know where this story leads. FDR, of course, won big that year. Final returns indicated that the Digest had overestimated Landon's popularity by as much as 19 points.

That same year, George Gallup (yes, that Gallup) performed one of his first nationwide polls, interviewing only about 1,500 people. He rightly predicted the outcome. How is it that a survey that polled so many more people performed so much worse? Well, it turns out that the Digest polled voters listed on automobile registries and in telephone books. This being the middle of the Great Depression, both cars and phones were luxury items, so only the very rich (and hence, very Republican-prone) were likely to have them. Gallup, meanwhile, polled a random sample of the population, period. Rich, poor, whatever. So while he only had a few people in his sample, he ended up with a much more accurate snapshot of the population.

This is the same kind of sampling that Nielsen does. Your comment about the Nielsen box not picking up on what TiVo may be doing nearby is fair, but discounting Nielsen ratings simply because they have a relatively small sample is like saying, "I don't trust my blood test because they didn't drain me dry and test all of it." Random samples, in fact, give us the best picture of what the public is thinking/feeling/doing at any given time, within a known margin of error (a little less than 3% for samples of 1,500 people).

Furthermore, TiVo owners, I would submit, are much like the auto and phone owners of 1936: they'll tend to skew the information towards upperclass viewers because only the rich (and the TV-dedicated) own them, as of now. So it doesn't matter that they all get surveyed; the data is still not representative of what everyone in the country, whether they own a TiVO box or not, is watching. One could make an argument that these are exactly the people the networks should be trying to reach (greater disposable income and all that), but not that the Nielsens are somehow "wrong" because they don't sample everyone in America.

Of course, I may have completely misinterpreted what you were trying to say. If so, I apologize for going off. As I said, it's a button... :)

--th

[ Next Thread | Previous Thread | Next Message | Previous Message ]
[ Contact Forum Admin ]

Forum timezone: GMT-8
VF Version: 3.00b, ConfDB:
Before posting please read our privacy policy.
VoyForums(tm) is a Free Service from Voyager Info-Systems.
Copyright © 1998-2019 Voyager Info-Systems. All Rights Reserved.