TiVo Community Forums Archives - Are ratings accurate?

TiVoCommunity.com
(c)opyright 1995-2005 All rights reserved
indexcheckTC
This area is a static history of posts in the TiVo Community Forum Archive.
This archive history was made for the simple indexing of search sites like Google.

Pages:1

Are ratings accurate?
(Click here to view the original thread with full colors/images)

Posted by: Family

This has bothered me. Are these television ratings accurate enough? We know that they're pretty consistant and are probably within range, but it seems that it only takes only a few hundred thousand more or less viewers within a whole nation to make a difference on success or cancellation of a show.

In politics there is eventually factual information that lets the pollsters/exit polls know if their system is working. For example, the presidential election told the networks that their exit polls needed a closer look. In television do these Nielsen people use the same criteria as the 1950's? How do they know wheither they are falling off track with regard to accuracy? Really how would they ever know how close they really are?

Do you think a poor outdated system of rating television shows is running the whole industry? I would think the actual advertisers or agencies would have created their own system of rating wheither a show was worthy for their budget. Doesn't the current method of sweeps/rating seem a little.... behind the times?

Posted by: tonyoci

I have never understood how they can be accurate. No one has ever asked me what I watch and I live in a strong target market (location, age, income etc.). However on the other hand the ratings do often track closely to my opinion, a top rated show is good and I like it, I start to think it's losing it's way and the ratings drop etc.

Of course they do not factor in Tivo'ers or anyone who records shows, that's a small market but since ratings are driven towards advertising (which we do not watch) that seems reasonable.

T

Posted by: whitson77

Not a chance. 1% of 110 million. And their methods are suspect. Go read how they get their numbers at the Nielson website. If I was cool I would link it, but I am lazy.

Posted by: brahamt

This is completely idiotic of me, but for some reason I don't believe statisical method works on TV watching. I can barely predict what I will watch at any given time. Aside from which, there is no way to predict how many TVs will be in use at a given time.

Please don't jump all over me, as I mentioned at the top, this opinion is a little irrational.

Posted by: innocentfreak

First off they may have changed it since some of the articles I read once upon a time, but the articles mentioned they don't ever include college students living in dorms. I don't know the reason why, but I pulled the article one day in a google search of Nielsen Ratings or something similar. The other problem is they depend on people to report themselves. For example you get the booklet and sit down and fill it out immediately writing down your favorite shows whether you watched them that week or not.

There has to be a better way, but as long as they are left to do it as they want and no one comes up with a better way. It is similar to radio. They still use the booklets. Surely by now someone could come up with the technology to monitor what is being watched. I realize then you would have the privacy people up in arms, but there has to be a better more accurate way.

Posted by: jsmeeker

I was a Nielsen houshold for one week late last year. I was a booklet person.

Are they accurate?? Somone asked this same question in Coffee Talk. Somone responded stating that statistical sampling methods are proven to be accurate. But to network executives, they are golden, so that is really all that matters.

Posted by: Family

Again I wonder why the major advertisers don't take a more active role in deciding what shows meet its audience requirements. Beyond sports I never hear anything about demographics in ratings. Thje importance simply seems to be number of viewers and the market share of a show.

Posted by: jerobi

quote:
I pulled the article one day in a google search of Nielsen Ratings or something similar. The other problem is they depend on people to report themselves.

This may play into the ratings (I've done it for radio twice), but I also want to point out that true "Neilsen Families" have every TV in their house wired up with accurate tracking systems. It's not just all books.

Also, TiVo ratings currently represent less than one percent of the total ratings.

Credit both to an article I read a few hours ago.

Posted by: jsmeeker

quote:
Originally posted by Family
Again I wonder why the major advertisers don't take a more active role in deciding what shows meet its audience requirements. Beyond sports I never hear anything about demographics in ratings. Thje importance simply seems to be number of viewers and the market share of a show.

There are demographics in ratings. You just probably don't normally see them. Typically, you just get the rating and share for all viewers, but Nielsen *does* track the ratings by various demogrpahic groups.

Posted by: KLB

I believe there are something like 5000 people that are the electronic 'neilsen families' that make up the overnights and majority of the ratings reportings and then in each of the major markets there are more but they aren't included as quickly.

5000????

The neilsens have been antiquated since the 70's and inaccurate to the maximum--- but none of the networks want to do anything about.

They don't want to wake up tomorrow and find out that only about one quarter of the people they though were watching Friends actually were watching.

Posted by: Mr. Funny Pants

quote:
Originally posted by KLB
The neilsens have been antiquated since the 70's and inaccurate to the maximum--- but none of the networks want to do anything about.

Source?

Posted by: lordsutch

A sample size of 5000 is valid for the whole United States; however, local ratings outside the top 10 markets or so would be wildly inaccurate based on such a small sample (which is why they use the diaries during sweeps). By comparison, most national telephone surveys only interview 1500 people or so.

Now, their selection methodology for the 5000 households may be suspect (since they omit college students, that may introduce some bias, especially since college students are in the core 18-49 demo). But the n (sample size) isn't a problem, in and of itself, at least for national TV and cable/satellite ratings.

Posted by: omv

5000 seems like too small a sample size to me, especially for more niche programs. If fewer people are watching a particular program, doesn't the error grow since 1 or 2 people can make a big impact? Well watched programs should be fairly accurate.

I need to bone up on my probablity for this case, as i'm assuming the deviation is something like sqrt(n), so for small n deviation is big. Certainly I know that holds for semiconductor implantation.

(To clarify, n is the number of viewers who watched, not the total 5000 pool - so a show with 100 viewers would have an error of 10, or 10%)

Posted by: rbird

I think Nielsen is a big bunch of scam artists. In addition to their (questionable at best) "ratings" service, now they have even leveraged their ratings monopoly into total control of what channels are available in satellite local-to-local packages (now codified into law via SHVIA!). It's unbelievable, but you have to give them credit for duping so many people for so long.

Bob