I understand that sampling is the way that these companies estimate the number of tv’s tuned to a particular program, but how is that number translated into total viewership? Do they multiplay that by some overall number derived from average people per household, average number of people in a bar, average number of people over at Howard’s house, etc? Because I see that an estimated 148.3 million viewers watched all or part of the Superbowl game. Where does the actual number come from, once the number of sets tuned to a station is determined? In particular, since the total number of people voting in 2004 was 126 million, how does this compare with the number of people of voting age who watched the game? thanks, Dopers. xo, C.
I can’t answer directly, but here are a couple threads on the topic:
Television viewer numbers: how do they know?
I wonder how DVR/Tivo is fitting into the mix, and why cable providers don’t include polling your set-top box as part of the service agreement. I’d think Time Warner or other massive provider would clean up.
Supposedly, because they don’t count DVR viewers as viewers, since they usually skip commercials. The Sci-Fi channel complains about this occasionally, since they have a particularly geek-heavy audience.
Another thing that Nielsen has started counting (or will start counting shortly) is viewers in bars and college dorms. Not sure how they will do that (and it may be addressed in Rhythmdvl links), but it appears that a significant amount of viewers have been left by the wayside.
As I read those links, there’s not attention paid to those things. That’s really my question. How do households, or other viewing venues, translate to individual viewers, and to what extent are those other venues even considered?
I thought Nielsen did count DVR viewings.
You seem to be correct–apparently they’ve been doing it a couple years now (separately from the “live viewer” numbers). Google “Nielsen DVR” though – nobody seems to like the numbers, or at least didn’t when they first came out. Several companies said they wouldn’t use them because of the ad-skipping.
Seems you can’t make everybody happy. Who’d have guessed?
Interesting to watch the percentage of DVR viewers climb, though (from that same search): it was about 5% in 2005, and about 17% in 2007, and as much as 40% in households above median income.