Impeach Trump!

[kalm]

That'd be suicide. The Democrats should've slow-walked this. They should've gone after Bolton and Mulvaney and gotten them on the record.

All so the Republicans can acquit.

I'm actually more worried about the precedent that Dershowtiz' argument will set. IF you think having a foreign government intervene in our election against your opponent is in the best interest of the US then it's perfectly fine. That's some dangerous precedent setting there, IMO.

I don't believe Dershowitz defended the Dems' attempts to get dirt on Trump from Ukraine, or pay for a dossier full of information collected from Russia and written by a former British intelligence officer.

Or is that not what you are talking about?

Yeah! And I’ll bet they used congressionally approved money too!

[kalm]

I don't know what you two guys were talking about but, actually, there is a point at which the sample size gives you the same "precision" (as a pratical matter) regardless of how much bigger the population gets.

In survey statistics there is a thing called the "finite population correction. You can read about it here: https://www.chegg.com/homework-help/def ... rection-31.

Let's say your sample size is 1000 and you're doing a survey to estimate the percent in a population of voters. If you assume 50% is the percent (proportion 0.5) the standard error if you assume infinite population is 0.015811388 (1.58%). Suppose the sample comes from a population of 1,000,000 voters. The finite population correction is 0.999000999 so the standard error goes to 0.015795593. If it goes to 100,000,000 voters the finite population correction is 0.99999001 and the standard error is 0.01581123.

So you can see that if you've got a sample size of 1000 you get practically the same standard error whether it's 1,000,0000 or 100,000,000 or infinity. As long as the population size is way bigger than the sample size a given sample size is about as good. The rule of thumb in a survey sampling class I took is that you reach that point when the sample size is <10% of the population size. The linked article says 5%.

But, for example, a sample size of 1000 is essentially just as good for a population of a billion or even infinity as it is for a population of a million or a hundred thousand. Even if the population of 10,000 the standard error for the example above would be 0.014231673.

So really, once the population is large, the "effectiveness" of a given sample size doesn't change much as it gets even larger and larger into infinity.

I agree that effectiveness isn't linear and after a certain point, the confidence level of your data is pretty high no matter the number (but will still be higher the larger the sample is, however only logarithmically). But isn't "finite population control" used as a factor in the sample size equation? As, if you want your effectiveness to be 90% (or whatever), you'd first do the equation above (and maybe another one or two) and then plug the results into the equation to get the sample size necessary to get that 90% confidence?

I could certainly be remembering wrong; it's been a million years since I took statistics. Ultimately my question is: does 435 reps accurately represent the US population with high confidence?

You seem to have a passion with this, so I trust your answer.

Jesus Brush Hogging Christ!

We’re not talking about the effect JSO’s poo has on downstream Black crappie populations in Lacassine Refuge.

Couple of big brained Bradley’s...

[CitadelGrad]

I don't believe Dershowitz defended the Dems' attempts to get dirt on Trump from Ukraine, or pay for a dossier full of information collected from Russia and written by a former British intelligence officer.

Or is that not what you are talking about?

Yeah! And I’ll bet they used congressionally approved money too!

Ukraine got it's congressionally authorized money, so I don't really see what you're complaining about.

I never knew you were so dedicated to the independence and sovereignty of Ukraine. I bet you had a cow when Obama refused to send arms to Kiev.

[CitadelGrad]

Also, Schiff refused to allow witnesses called by the GOP.

False. Volker, Hale, and Morrison were on the GOP witness list.

https://www.foxnews.com/politics/read-h ... t-hearings

Did those four names comprise the entire list? No.

Fuck you.

[Gil Dobie]

Just glad California didn't get it's way again. Cali is the only reason the election was close in the electoral college and Hillary winning the popular vote. I can see Calexit getting a little momentum again in the news.

[Ibanez]

Wow. The Reps are just lying now. They’ve named all the witnesses that were interviewed and said they didn’t get to cross examine.


Then who were those Republicans questioning Hill, Sondland, Vindman, et al...?


Sent from my iPhone using Tapatalk

There were 17 witnesses. The GOP was unable to depose or cross examine 11 of them, if my recollection is correct.

Also, Schiff refused to allow witnesses called by the GOP.

Are you referring to the dispositions with Republicans present? Or the other interviews in which the White House refused to be part of? Impeachment hearings aren’t criminal trials. Don’t confuse the two.


Sent from my iPhone using Tapatalk

[kalm]

Yeah! And I’ll bet they used congressionally approved money too!

Ukraine got it's congressionally authorized money, so I don't really see what you're complaining about.

I never knew you were so dedicated to the independence and sovereignty of Ukraine. I bet you had a cow when Obama refused to send arms to Kiev.

I can see you haven’t followed this story much so I won’t call you a dolt.

I could give a frogs fat ass if the Republicans would like to investigate the Dems. It’s fairly obvious that the Hunter Biden thing is corrupt on at least some level...so fire away! The Steele Dossier work may be illegal too but it’s clearly different than what Trump was charged with.

Crawl back under your bridge and do some research.

[Skjellyfetti]

False. Volker, Hale, and Morrison were on the GOP witness list.

https://www.foxnews.com/politics/read-h ... t-hearings

Did those four names comprise the entire list? No.

Fuck you.

You doin ok, bud?

[AZGrizFan]

I don't know what you two guys were talking about but, actually, there is a point at which the sample size gives you the same "precision" (as a pratical matter) regardless of how much bigger the population gets.

In survey statistics there is a thing called the "finite population correction. You can read about it here: https://www.chegg.com/homework-help/def ... rection-31.

Let's say your sample size is 1000 and you're doing a survey to estimate the percent in a population of voters. If you assume 50% is the percent (proportion 0.5) the standard error if you assume infinite population is 0.015811388 (1.58%). Suppose the sample comes from a population of 1,000,000 voters. The finite population correction is 0.999000999 so the standard error goes to 0.015795593. If it goes to 100,000,000 voters the finite population correction is 0.99999001 and the standard error is 0.01581123.

So you can see that if you've got a sample size of 1000 you get practically the same standard error whether it's 1,000,0000 or 100,000,000 or infinity. As long as the population size is way bigger than the sample size a given sample size is about as good. The rule of thumb in a survey sampling class I took is that you reach that point when the sample size is <10% of the population size. The linked article says 5%.

But, for example, a sample size of 1000 is essentially just as good for a population of a billion or even infinity as it is for a population of a million or a hundred thousand. Even if the population of 10,000 the standard error for the example above would be 0.014231673.

So really, once the population is large, the "effectiveness" of a given sample size doesn't change much as it gets even larger and larger into infinity.

I agree that effectiveness isn't linear and after a certain point, the confidence level of your data is pretty high no matter the number (but will still be higher the larger the sample is, however only logarithmically). But isn't "finite population control" used as a factor in the sample size equation? As, if you want your effectiveness to be 90% (or whatever), you'd first do the equation above (and maybe another one or two) and then plug the results into the equation to get the sample size necessary to get that 90% confidence?

I could certainly be remembering wrong; it's been a million years since I took statistics. Ultimately my question is: does 435 reps accurately represent the US population with high confidence?

You seem to have a passion with this, so I trust your answer.

Let’s put it this way, trip, your math and statistics struggles aside: we struggle (as a country) to find 435 people who are capable of managing a marginally functioning government process. How in the flying FUCK do you propose we find TEN THOUSAND people to do that, when we can’t hardly find 435?

[kalm]

I agree that effectiveness isn't linear and after a certain point, the confidence level of your data is pretty high no matter the number (but will still be higher the larger the sample is, however only logarithmically). But isn't "finite population control" used as a factor in the sample size equation? As, if you want your effectiveness to be 90% (or whatever), you'd first do the equation above (and maybe another one or two) and then plug the results into the equation to get the sample size necessary to get that 90% confidence?

I could certainly be remembering wrong; it's been a million years since I took statistics. Ultimately my question is: does 435 reps accurately represent the US population with high confidence?

You seem to have a passion with this, so I trust your answer.

Let’s put it this way, trip, your math and statistics struggles aside: we struggle (as a country) to find 435 people who are capable of managing a marginally functioning government process. How in the flying FUCK do you propose we find TEN THOUSAND people to do that, when we can’t hardly find 435?

And there’s already 40,000 lobbyists to help them now. We’d need 800,000 lobbyists with Trips plan.

[JohnStOnge]

I could certainly be remembering wrong; it's been a million years since I took statistics. Ultimately my question is: does 435 reps accurately represent the US population with high confidence?

You seem to have a passion with this, so I trust your answer.

I normally don't like to talk about why people should believe me but in this case I will say that I did have a graduate level survey statistics course and have done a number of statistical surveys. I won a "best presentation" award at an American Fisheries Society meeting for a presentation on one of them. Not public opinion surveys. Biological and environmental things. The one I won the best presentation award for had to do with estimating sizes, species composition, and sex ratios for a certain fishery. But the statistical principles are the same. It's just that an animal selected during a survey can't refuse to let you measure it and determine what species and sex it is. Pollsters have the daunting complication of dealing with human beings who can avoid being "measured" if selected and, as some would say, lie to them.

The way to look at it is this: Say you are estimating the percentage in a population with characteristic X and you are going to select a simple random sample of the population. You assume 50 percent because 50 percent is the point at which the standard error and the margin of error based on that is largest.

You ask yourself what the margin of error will be if you get a random sample of 435. With an infinite population the answer is that the at 435 the margin of error will be + or - 4.7 percentage points.

So that's for an infinite population. How much difference does it make if the population is a city of 100,000 or a country of 300,000,000? The answer, as a practical matter, is "none." if you're going to the nearest 10th of a percentage point the margin of error is 4.7 if the population your sampling is 100,000, 300,000,000, or infinity.

Now, a bigger sample will decrease the margin of error but there is a principle of diminishing returns assocciated with that. The image below shows what happens with the margin of error while assuming 50% and infinite population as you start at a sample size of 100 then increase the sample size in increments of 100 until you get to 3,000. The Y axis is margin of error and the X axis is sample size. You can see that at first the incremental increases make a notable difference but the gain in terms of reducing the margin of error gets smaller and smaller as the sample sizes gets larger. You can see you have to make a decision about where to stop based on how much you're going to gain by increasing the sample size. I would say you'd want to go farther than 435 if you can.

[AZGrizFan]

Let’s put it this way, trip, your math and statistics struggles aside: we struggle (as a country) to find 435 people who are capable of managing a marginally functioning government process. How in the flying FUCK do you propose we find TEN THOUSAND people to do that, when we can’t hardly find 435?

And there’s already 40,000 lobbyists to help them now. We’d need 800,000 lobbyists with Trips plan.

[JohnStOnge]

[we struggle (as a country) to find 435 people who are capable of managing a marginally functioning government process.

It's a lot more than 435 people but I think that when you step back and look at the big picture the United States government does a very good job. Just try to imagine what would happen if you could snap your fingers and everything the United States government does stopped. It would be complete chaos.

Yes there are plenty of problems and one can find plenty of examples of incompetence in certain places. But when you look at the big picture what our government does to maintain order and create a background against which things can function is very impressive.

Managing a population of 330 million people as well as all of the resource considerations associated with the United States is not easy.

[CID1990]

I could certainly be remembering wrong; it's been a million years since I took statistics. Ultimately my question is: does 435 reps accurately represent the US population with high confidence?

You seem to have a passion with this, so I trust your answer.

I normally don't like to talk about why people should believe me but in this case I will say that I did have a graduate level survey statistics course and have done a number of statistical surveys. I won a "best presentation" award at an American Fisheries Society meeting for a presentation on one of them. Not public opinion surveys. Biological and environmental things. The one I won the best presentation award for had to do with estimating sizes, species composition, and sex ratios for a certain fishery. But the statistical principles are the same. It's just that an animal selected during a survey can't refuse to let you measure it and determine what species and sex it is. Pollsters have the daunting complication of dealing with human beings who can avoid being "measured" if selected and, as some would say, lie to them.

The way to look at it is this: Say you are estimating the percentage in a population with characteristic X and you are going to select a simple random sample of the population. You assume 50 percent because 50 percent is the point at which the standard error and the margin of error based on that is largest.

You ask yourself what the margin of error will be if you get a random sample of 435. With an infinite population the answer is that the at 435 the margin of error will be + or - 4.7 percentage points.

So that's for an infinite population. How much difference does it make if the population is a city of 100,000 or a country of 300,000,000? The answer, as a practical matter, is "none." if you're going to the nearest 10th of a percentage point the margin of error is 4.7 if the population your sampling is 100,000, 300,000,000, or infinity.

Now, a bigger sample will decrease the margin of error but there is a principle of diminishing returns assocciated with that. The image below shows what happens with the margin of error while assuming 50% and infinite population as you start at a sample size of 100 then increase the sample size in increments of 100 until you get to 3,000. The Y axis is margin of error and the X axis is sample size. You can see that at first the incremental increases make a notable difference but the gain in terms of reducing the margin of error gets smaller and smaller as the sample sizes gets larger. You can see you have to make a decision about where to stop based on how much you're going to gain by increasing the sample size. I would say you'd want to go farther than 435 if you can.

It took like 6 upward swipes of my thumb to ignore this one


Sent from my iPhone using Tapatalk

[SunCoastBlueHen]

I normally don't like to talk about why people should believe me but in this case I will say that I did have a graduate level survey statistics course and have done a number of statistical surveys. I won a "best presentation" award at an American Fisheries Society meeting for a presentation on one of them. Not public opinion surveys. Biological and environmental things. The one I won the best presentation award for had to do with estimating sizes, species composition, and sex ratios for a certain fishery. But the statistical principles are the same. It's just that an animal selected during a survey can't refuse to let you measure it and determine what species and sex it is. Pollsters have the daunting complication of dealing with human beings who can avoid being "measured" if selected and, as some would say, lie to them.

The way to look at it is this: Say you are estimating the percentage in a population with characteristic X and you are going to select a simple random sample of the population. You assume 50 percent because 50 percent is the point at which the standard error and the margin of error based on that is largest.

You ask yourself what the margin of error will be if you get a random sample of 435. With an infinite population the answer is that the at 435 the margin of error will be + or - 4.7 percentage points.

So that's for an infinite population. How much difference does it make if the population is a city of 100,000 or a country of 300,000,000? The answer, as a practical matter, is "none." if you're going to the nearest 10th of a percentage point the margin of error is 4.7 if the population your sampling is 100,000, 300,000,000, or infinity.

Now, a bigger sample will decrease the margin of error but there is a principle of diminishing returns assocciated with that. The image below shows what happens with the margin of error while assuming 50% and infinite population as you start at a sample size of 100 then increase the sample size in increments of 100 until you get to 3,000. The Y axis is margin of error and the X axis is sample size. You can see that at first the incremental increases make a notable difference but the gain in terms of reducing the margin of error gets smaller and smaller as the sample sizes gets larger. You can see you have to make a decision about where to stop based on how much you're going to gain by increasing the sample size. I would say you'd want to go farther than 435 if you can.

It took like 6 upward swipes of my thumb to ignore this one


Sent from my iPhone using Tapatalk

It only took me four. You're a sub-par swiper.

Sent from my SM-J810Y using Tapatalk

[AZGrizFan]

[we struggle (as a country) to find 435 people who are capable of managing a marginally functioning government process.

It's a lot more than 435 people but I think that when you step back and look at the big picture the United States government does a very good job. Just try to imagine what would happen if you could snap your fingers and everything the United States government does stopped. It would be complete chaos.

Yes there are plenty of problems and one can find plenty of examples of incompetence in certain places. But when you look at the big picture what our government does to maintain order and create a background against which things can function is very impressive.

Managing a population of 330 million people as well as all of the resource considerations associated with the United States is not easy.

The fact you EVER called yourself a conservative is fucking comical.

to sit here and justify the size and expanse of the government, like we’d all be unable to function without their benevolent oversight. WAFJ.

And besides the 435 were REPRESENTATIVES, specifically. Trip wants 10,000 when we can’t even find 435 competent ones.

[∞∞∞]

I could certainly be remembering wrong; it's been a million years since I took statistics. Ultimately my question is: does 435 reps accurately represent the US population with high confidence?

You seem to have a passion with this, so I trust your answer.

I normally don't like to talk about why people should believe me but in this case I will say that I did have a graduate level survey statistics course and have done a number of statistical surveys. I won a "best presentation" award at an American Fisheries Society meeting for a presentation on one of them. Not public opinion surveys. Biological and environmental things. The one I won the best presentation award for had to do with estimating sizes, species composition, and sex ratios for a certain fishery. But the statistical principles are the same. It's just that an animal selected during a survey can't refuse to let you measure it and determine what species and sex it is. Pollsters have the daunting complication of dealing with human beings who can avoid being "measured" if selected and, as some would say, lie to them.

The way to look at it is this: Say you are estimating the percentage in a population with characteristic X and you are going to select a simple random sample of the population. You assume 50 percent because 50 percent is the point at which the standard error and the margin of error based on that is largest.

You ask yourself what the margin of error will be if you get a random sample of 435. With an infinite population the answer is that the at 435 the margin of error will be + or - 4.7 percentage points.

So that's for an infinite population. How much difference does it make if the population is a city of 100,000 or a country of 300,000,000? The answer, as a practical matter, is "none." if you're going to the nearest 10th of a percentage point the margin of error is 4.7 if the population your sampling is 100,000, 300,000,000, or infinity.

Now, a bigger sample will decrease the margin of error but there is a principle of diminishing returns assocciated with that. The image below shows what happens with the margin of error while assuming 50% and infinite population as you start at a sample size of 100 then increase the sample size in increments of 100 until you get to 3,000. The Y axis is margin of error and the X axis is sample size. You can see that at first the incremental increases make a notable difference but the gain in terms of reducing the margin of error gets smaller and smaller as the sample sizes gets larger. You can see you have to make a decision about where to stop based on how much you're going to gain by increasing the sample size. I would say you'd want to go farther than 435 if you can.

Good stuff JSO, thank you.

[houndawg]

That'd be suicide. The Democrats should've slow-walked this. They should've gone after Bolton and Mulvaney and gotten them on the record.

All so the Republicans can acquit.

I'm actually more worried about the precedent that Dershowtiz' argument will set. IF you think having a foreign government intervene in our election against your opponent is in the best interest of the US then it's perfectly fine. That's some dangerous precedent setting there, IMO.

I don't believe Dershowitz defended the Dems' attempts to get dirt on Trump from Ukraine, or pay for a dossier full of information collected from Russia and written by a former British intelligence officer.

Or is that not what you are talking about?

The only question relevant to Dershowitz is: Boxers or briefs?

[CitadelGrad]

There were 17 witnesses. The GOP was unable to depose or cross examine 11 of them, if my recollection is correct.

Also, Schiff refused to allow witnesses called by the GOP.

Are you referring to the dispositions with Republicans present? Or the other interviews in which the White House refused to be part of? Impeachment hearings aren’t criminal trials. Don’t confuse the two.


Sent from my iPhone using Tapatalk

It's depositions, not dispositions. Anyway, why would only the Dems get to depose them?

[CitadelGrad]

Ukraine got it's congressionally authorized money, so I don't really see what you're complaining about.

I never knew you were so dedicated to the independence and sovereignty of Ukraine. I bet you had a cow when Obama refused to send arms to Kiev.

I can see you haven’t followed this story much so I won’t call you a dolt.

I could give a frogs fat ass if the Republicans would like to investigate the Dems. It’s fairly obvious that the Hunter Biden thing is corrupt on at least some level...so fire away! The Steele Dossier work may be illegal too but it’s clearly different than what Trump was charged with.

Crawl back under your bridge and do some research.

I don't believe I mentioned the Steele Dossier, which was paid for by Dems and contained information obtained from Russia by a former British MI-6 officer.

[CitadelGrad]

Did those four names comprise the entire list? No.

Fuck you.

You doin ok, bud?

You didn't answer the question.

Fuck you again.

[Ibanez]

Are you referring to the dispositions with Republicans present? Or the other interviews in which the White House refused to be part of? Impeachment hearings aren’t criminal trials. Don’t confuse the two.


Sent from my iPhone using Tapatalk

It's depositions, not dispositions. Anyway, why would only the Dems get to depose them?

Do subjects of investigations participate every step of the way into investigations? I'm unaware of someone, let's say a murderer, participating in the discovery phase with the prosecution. I know the prosecutor has to share evidence (which the House did).

The White House had an opportunity and they declined. Then they did participate at the very end. Furthermore, Republicans on the committees were involved.

[Ibanez]

Yeah! And I’ll bet they used congressionally approved money too!

Ukraine got it's congressionally authorized money, so I don't really see what you're complaining about.

I never knew you were so dedicated to the independence and sovereignty of Ukraine. I bet you had a cow when Obama refused to send arms to Kiev.

Yes they did. After the DoD TWICE certified that they had done enough to combat corruption.

So why was Trump withholding aid?

Oh, that's right. He was waiting for them to announce investigations into his political opponent. Tell me again how investigating Biden is critical to the national security and is in the national interest. So far, I dont know the answer.

Also, saying they eventually got it as an excuse to withhold is BS.

[houndawg]

It's a lot more than 435 people but I think that when you step back and look at the big picture the United States government does a very good job. Just try to imagine what would happen if you could snap your fingers and everything the United States government does stopped. It would be complete chaos.

Yes there are plenty of problems and one can find plenty of examples of incompetence in certain places. But when you look at the big picture what our government does to maintain order and create a background against which things can function is very impressive.

Managing a population of 330 million people as well as all of the resource considerations associated with the United States is not easy.

The fact you EVER called yourself a conservative is fucking comical.

to sit here and justify the size and expanse of the government, like we’d all be unable to function without their benevolent oversight. WAFJ.

And besides the 435 were REPRESENTATIVES, specifically. Trip wants 10,000 when we can’t even find 435 competent ones.

You gotta admit that ol' Johnny trolled for years a site that isn't easy to troll.

[houndawg]

You doin ok, bud?

You didn't answer the question.

Fuck you again.

I laugh my ass off thinking of you stuck in St.Louis. Your company must hate you.