| BMaker11 said: I understand that. That's just how statistics work. Hell, even at the 100K level that was proposed, you *could* get 100K partisan people (not out of the question since these consoles sells 200K a week). But if I've learned anything from my statistics courses, anything above 2000 is asking too much. It takes too much time and effort (and money) to gather that many samples, and calculate accordingly. However, most of the time, at around 2000, it's damned accurate. But, yes, I do understand that any time you do random sampling, there's the potential for random sampling error. Unless you can sample an entire population ("sample" isn't the correct word when referring to a population, I know) you will never always be automatically correct. But then again, those high are numbers are simply not feasible |
I agree that the number sampled is valid. I am not trying to attack their methodology on this one. The results are just an anomaly that eventually happens. It is the only logical explanation if we assume they correctly performed their job. It could be that they intentionally led people to choose certain options with how the questions were phrased, but it seems unlikely.
Although it actually might be just that they phrased it more as a wish list than an intent to buy. The end justification then being the items that end up lower than expected would be priced higher than the general consumer is willing to pay. I am reluctant to believe any of the items are still viewed as over priced right now save for perhaps the iPad.
