PictishBeast wrote:It's like, math and shit. You'd think geeks would be all over it.
My point wasn't focusing on the small aspect, which you are right the sample size is mathematically enough, but it's the inclusion part that's basically the bane of any research or polling anywhere. No one can be sure what demographics or areas were chosen (unless this is told to each station that requests information? not exactly clear) and always leaves potential to leave out certain audiences because of this. Normally, just is just how statistics works because you can only study what is in front of you, the literal point being "Well lets take a cut of the population since we statstically don't need a large size". But the point I was comparing to was to Twitter/online spaces which not only has a much bigger sample pool to draw from but also isn't quite as likely to select w/r/t area demographics or the like. Basically, it's at the bottom survey system that you posted:
How many people are there in the group your sample represents?...This is not a problem. The mathematics of probability proves the size of the population is irrelevant unless the size of the sample exceeds a few percent of the total population you are examining. This means that a sample of 500 people is equally useful in examining the opinions of a state of 15,000,000 as it would a city of 100,000.
The confidence interval calculations assume you have a genuine random sample of the relevant population. If your sample is not truly random, you cannot rely on the intervals. Non-random samples usually result from some flaw in the sampling procedure. An example of such a flaw is to only call people during the day and miss almost everyone who works. For most purposes, the non-working population cannot be assumed to accurately represent the entire (working and non-working) population.
My point being that it's (arguably) easier to dismiss flaws when taking samples from internet bases now than TV ones that have the base selected since it's easier to avoid situations like the above thanks to Twitter/mobile/online usage being so prevalent that it has members from all/most demographics. Still puts kids <12 at a weird spot though.
The other half of the argument here is the slow decline of TV relevance. In my area, which I know is anecdotal since I'm in college, almost everyone around me has a Twitter but not even half of them watch TV anymore. This is what prompted the conversation, because there is
potentially a flaw in the TV base or how the American Population base relates to TVs, which would make these ratings not as reliable depending on who exactly (Hasbro in our case) is being targeted, which is what we mean by "Outdated".
It has its uses as I pointed out, and you can make further arguments about how sample sizes/statistics work or how they would be the same for online spaces too. I'm just making the point that there IS room to look at data from mobile/internet and that looking into data from there is the future while TV data is slowly becoming more and more irrelevant.
PictishBeast wrote:This is true but Nielsen is less concerned with figuring out the total viewership of the show and more concerned with figuring out the viewership of the
ads in the show (since that's where most of the money is coming in to the network so that they can continue to produce new episodes).
Digital episodes of a TV show are sold separately and can be measured directly by # of impressions or # of plays using server data. There's no need to rely on indirect statistical sampling like Nielsen and there's not really much of a point in lumping the two numbers together since they're essentially two different products with two different revenue streams.
This is basically what I meant with that next sentence from the posted quote; to us viewers, we only care about the raw number of plays versus the Companies who want to track demographics and ads therein. There is a chance you can do both forms of tracking with Digital episodes, but that'd require a different system in place.
Wylie wrote:Statistics are weird and counterintuitive and they lie a lot.
While I agree with this sentence, the example you bring up is kind of silly. Are families really traveling out much in this economy?
