War in Israel

[Mirage]

"Methodology: The survey of 1,429 adults was conducted Nov. 6-9 by the Marist Poll by phone, both cell phones and landlines using live interviewers, by text or online in both English and Spanish. The margin of error is + or - 3.4 percentage points".

I'm curious how they arrived at a 3.4% margin of error by sampling 1,429/330,000,000 Americans ?

Statistics and math.

 

[Average Monke]

Statistics and math.

Impossible...

 

[sevenhelmet]

Lies, damn lies, and…

 

[BattlingTrain]

It was 4.2069% by my math but I was in the Army so I don't count too good

 

[Griz882]

Impossible...

I think you men unpossible.

 

[nittany03]

"Methodology: The survey of 1,429 adults was conducted Nov. 6-9 by the Marist Poll by phone, both cell phones and landlines using live interviewers, by text or online in both English and Spanish. The margin of error is + or - 3.4 percentage points".

I'm curious how they arrived at a 3.4% margin of error by sampling 1,429/330,000,000 Americans ?

Because if you took Stat 101, you would learn that as long as the demographics of those 1,429 people are a close match to the demographics of 330 million Americans, then you will get +- 3.4% of what you would have gotten by polling all 330 million . . . and it's much cheaper to do. The hard part is getting an accurate sample population.

 

[croakerfish]

The hard part is getting an accurate sample population.

Out, liars!

 

[FLGUY]

Because if you took Stat 101, you would learn that as long as the demographics of those 1,429 people are a close match to the demographics of 330 million Americans, then you will get +- 3.4% of what you would have gotten by polling all 330 million . . . and it's much cheaper to do. The hard part is getting an accurate sample population.

Not trying to go on a tangent, just curious: Wouldn’t something like location (among other variables) play a significant part in the results of a random poll? Like if you were to poll the same sample population that matches the required demographics in say, California and then did the same in Alabama, would you not get a significant difference in results?

 

[Mirage]

Not trying to go on a tangent, just curious: Wouldn’t something like location (among other variables) play a significant part in the results of a random poll? Like if you were to poll the same sample population that matches the required demographics in say, California and then did the same in Alabama, would you not get a significant difference in results?

The idea is to poll a random sampling of the target demographic. If it's a national poll, for the results to be valid you must poll a random cross section of the entire country. Your last sentence is an oxymoron, because the required demographic by definition cannot reside just in California or Alabama.

 

[Brett327]

Not trying to go on a tangent, just curious: Wouldn’t something like location (among other variables) play a significant part in the results of a random poll? Like if you were to poll the same sample population that matches the required demographics in say, California and then did the same in Alabama, would you not get a significant difference in results?

Yes. There's an entire field of political science devoted to stuff like this.

 

[FLGUY]

The idea is to poll a random sampling of the target demographic. If it's a national poll, for the results to be valid you must poll a random cross section of the entire country. Your last sentence is an oxymoron, because the required demographic by definition cannot reside just in California or Alabama.

My statement is not an Oxymoron, as I think it’s safe to say that not every poll that is conducted is a national poll. Interesting stuff nonetheless.

 

[ChuckMK23]

I get called regularly (3-4 times per year) on Marist national polls - I can't figure out how - possibly the periodicals I subscribe to (Scientific American, MIT Technology, Harpers, NYT...) - inevitably the calls come when I am on the commode! Their questions are thoughtfully constructed and have a conversational tone - and you have the ability to ask the pollster for clarification on questions you are being polled on. I always find it an enjoyable experience - and feeling like I am adding something to the dialog.

 

[sevenhelmet]

I get called regularly (3-4 times per year) on Marist national polls - I can't figure out how - possibly the periodicals I subscribe to (Scientific American, MIT Technology, Harpers, NYT...) - inevitably the calls come when I am on the commode! Their questions are thoughtfully constructed and have a conversational tone - and you have the ability to ask the pollster for clarification on questions you are being polled on. I always find it an enjoyable experience - and feeling like I am adding something to the dialog.

Sounds better than the standard machine disconnect or heavy breathing followed by “we have been trying to reach you for your vehicle’s extended warranty…” that I typically get.

In all seriousness, any thoughts on these interviews being conducted by phone conversation possibly skewing the results? Not maligning anyone, but it’s my observation that only certain demographics are likely to pick up the phone and have a conversation with a stranger.

 

[ChuckMK23]

Sounds better than the standard machine disconnect or heavy breathing followed by “we have been trying to reach you for your vehicle’s extended warranty…” that I typically get.

In all seriousness, any thoughts on these interviews being conducted by phone conversation possibly skewing the results? Not maligning anyone, but it’s my observation that only certain demographics are likely to pick up the phone and have a conversation with a stranger.

Their caller ID always says "Marist Institute for Public Opinion"....

 

[sevenhelmet]

Their caller ID always says "Marist Institute for Public Opinion"....

You know scammers fake incoming numbers and caller IDs, right?

With how inundated with spam callers the general market is, I bet a lot of people still don’t pick up.