It certainly does say *something *about the larger group–we just can’t be sure how *accurately *it says it. However, the more responses we get, the more accurate it’s likely to be, and we’re up over 160 now. Perhaps more telling, the proportions of people at various levels of taxation seem to have stayed fairly consistent.
I actually voted for one category too high, because at first I was looking at my “taxes owed” without taking into account my tax credits, which dropped me into the lower category.
I guess that leaves you with the option of making less money! Or giving more of it away.
Right, and that lack of certainty about the accuracy effectively means it says nothing about any larger group.
I’m sure you’ll pardon me if I retain a certain degree of skepticism as to what may be influencing your opinion regarding the accuracy of the poll.
It’s because he doesn’t get enough hugs.
And I’m sure you’ll pardon me for having taken statistics in college. It’s interesting that you keep insisting that I am the biased one when my views on the accuracy of this poll is absolutely consistent with the objectiver scientific consensus on the matter. You may want to re-evaluate which of us is biased.
If enough data are collected, then it will be very representative of the larger group: all SDMB members.
Which is what **Shot From Guns **wanted in the first place.
Right, but only by brute force, and you’d need way more than a majority to respond to even get results that could even be close to accurate (further assuming that everyone interpreted the quesrion in the same manner, which we’ve already seen is not the case, and that everyone was truthful).
The reason that random sampling delivers results within a known margin of error is that we can know the odds that a random sample differs from the population. We can’t know those odds for a self-selected sample. So, it’s not even a matter of assuming a larger margin of error–we just can’t create a margin of error to begin with.
Now, if ALL members of a population self-select to answer a poll, then you’ve lucked out and the poll does say something about the population. With all of the different choices in this poll, wed have to have a huge percentage of SDMB members respond before being at all sure that the non-responders wouldn’t wildly change the results we’ve already gotten.
It’s not that difficult, people. Line 60 on Form 1040EZ.
TurboTax says my effective federal income tax rate was 9.7%.
I pulled up my tax return (Married filing together) in TurboTax and when I put my curser over my refund amount, it said I paid $534 in taxes, so picked the $500-$999 option.
Now, that is low now since my wife was unemployed for most of last year so only had to pay taxes on her unemployment insurance, and I’m in school so didn’t work much (plus my place of employment took FOREVER to give me my intern raise which doubled my pay)…
But, once I graduate in a little over a year, I’ll be in that $20,000+ bracket till I retire.
I find it interesting that you’re *completely *dismissing the poll as having any meaning at all. Even a “Hrm, that’s interesting” would suffice. Obviously there’s no way to know, but I suspect that if the poll had turned out as *you *expected, you’d be pointing to it in some way, shape, or form, statistically valid or no.
The fact that the proportions have stayed fairly consistent throughout, I think, is particularly interesting and perhaps suggestive that the numbers are a decently accurate reflection of the general population. Of course, it could just reflect the type of people most likely to come in and respond to a poll like this.
Wrong again, SFG. I wouldn’t say this poll had any validity whatsoever no matter what the results are. Again, you are showing your bias.
You may certainly continue to think so. I assure you that your opinion of me will continue to not affect me in the least.