Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
Why is anyone shocked? The list is worthless. DC up near the top while Maryland and Virginia are much lower ranked makes no sense. Most of the rude drivers in DC. are from those states. Does "DC" just mean the city itself or the Metro area, which is full or rude drivers, who are mostly from those states. Couple this obvious problem with the poll with all of the others mentioned and it's clearly a worthless poll. Why be shocked at that. The media constantly puts out worthless polls. It's just another day in America.
USA Today is a birdcage liner of a newspaper. They are about 1/2 step above the National Enquirer.
Did they do this survey on a Monday or a Friday? I notice people are the most miserable a holes on the road on mondays but come friday night rush hour everyone is happy to wave you on to go first. On monday they kindly will flip you the bird and cut you off.
I'm still waiting for Boxus to come back and explain how I'm an idiot for mocking a by state survey with only 2000 total people while real polling outfits use at least 500 in a single state.
Hmmmm gonna have to start lookin' under rocks I guess....lol
any high population density area has the potential to have rude drivers.. rude can mean many things to many people... around here a lot of people drive too defensively going 5 under for no real reason though other then that I cant say people are all that rude around here... am I the rude one for simply wanting them to do the speed limit?
Really? 2000 observations is enough to derive credible results comparing the drivers in 50 different states?
Here is how someone with an actual job producing credible results does a survey.
Here let me explain this as simple as I can to you, you are more than welcome to enroll in a college statistics class though.
200 million people in the US have a driver's license. With a confidence level of 95%, and a confidence interval of 5%, the sample size will be 385 people. That would be using the entire US as the pool size.
The methodology states;
"Insure.com commissioned a survey of 2,000 licensed drivers, half women and half men, with respondents representing all areas of the country according to Census population data. The state rankings were calculated using a ratio of the nationwide votes for drivers of the state divided by the number of respondents from the state. The survey was fielded in July 2014."
This shows me they used a proportionally, stratified sample based on gender and area. Since it is obvious they did not develop the sample pool from each area being measured, a statistically significant defensible measurement was not used, instead, a fair estimate is used for the results. Since the survey failed to provide more detailed methodology, and failed to properly declare itself as statistically significant or a fair estimate, I will agree with you that on the face of it, it is a poor survey. But that is not because of the 2000 people, which is more than enough for a statistically significant sample from 200 million, but because the methodology is not clear, and that are comparing state to state, though the sample pool was not derived from each state. Also, the methodology did not state it is a fair estimate, which is should have, instead, it implies it is a statistically significant measurement, which it is not.
I have no idea what this has to do with anything. The size of the sample pool has numerous factors in it. As the methodology in the article states, it has a confidence level of 95% with an interval of 1%, and is weighted according to demographics. the stratification that Gallup used obviously increased the sample size needed. You can go browse Rasmussen or Gallup and see they only survey 2000-3000 or so people for the presidential elections.
Quote:
Originally Posted by Mathguy
I don't expect you to show your face around here again after this exchange so appreciate the fact that I'm laughing at you as you demonstrated you complete lack of experience while lecturing me. I can only pray that if you are doing analysis for a living that it's for one of our competitors.
Why would I not show my face around here? You are the one that said 2000 people is not enough, I have clearly shown 2000 people is more than enough for a sample, unless you need me to break the math down for you.
Now, I do agree with you about the methodology, and more importantly, the lack of sufficient details in the methodology and the interpreted results from the methodology, but I do not agree that 2000 people cannot be a proper sample size.
As for working for your competitor, maybe, I work in the financial industry, maybe we even work for the same company.
Quote:
Originally Posted by Mathguy
Of course if they were asking 2000 people something homogenous like, "Do you like grapes" then yes, 2000 is probably enough. I'd be nicer in this exchange but you went full derp right out of the box so I don't see the need.
Well, I did not go full "derp' as you stated, I was responding to your post where you implied the sample size was not large enough, I disagreed. Perhaps you (and myself) should have elaborated more on the the specific area of the survey in which there was a disagreement, instead of the broad "2000 is not enough".
Quote:
Originally Posted by Mathguy
I'm still waiting for Boxus to come back and explain how I'm an idiot for mocking a by state survey with only 2000 total people while real polling outfits use at least 500 in a single state.
Hmmmm gonna have to start lookin' under rocks I guess....lol
How many people they use in the state depends on the number of people they are polling in the state, the stratification of the sample, what the sample is for (as in if it is a comparison, then the sample must be derived from comparison pools), etc. As I stated , the presidential elections uses only 2000 - 3000 people or so.
As a side note, a good read is Gallup's fall from grace due to the last few elections.
Here let me explain this as simple as I can to you, you are more than welcome to enroll in a college statistics class though.
.................
But that is not because of the 2000 people, which is more than enough for a statistically significant sample from 200 million, but because the methodology is not clear, and that are comparing state to state, though the sample pool was not derived from each state.
That was exactly my point....which you missed and then reamed me over.
2000 can be plenty if you are asking a question like "Do you approve of how Obama is doing as president?"
But to discern opinions about specific state driving rudeness?
Errrrr......nope.
Tell ya what, show me a poll showing how each states electoral votes are going to go....using just 2000 people. Not a national "hey how do you like Obama" but actually show me how someone could with just 2000 random samples from across the US they could determine how the electoral votes would result.
I'll wait.
My guess is that you know stats just fine, but suck at reading.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
