Do the Left & Right Argue Differently? - Page 2 - PeachParts Mercedes-Benz Forum

davidmash · #16 08-20-2011, 12:15 AM

Take a statistical analysis class. A 10 person sampling of such a huge population has a margin of error so large that the sample/analysis is useless.

elchivito · #17 08-20-2011, 12:17 AM

At the very least, you should rename your study. Call it something like "Do Those Who Self-Identify as Liberal or Conservative on This forum Argue Differently?" Generalizing it to anything beyond that is hardly scientific.

sjh · #18 08-20-2011, 12:22 AM

Quote:

Originally Posted by davidmash

Take a statistical analysis class. A 10 person sampling of such a huge population has a margin of error so large that the sample/analysis is useless.

I will use very small words.

I have worked with data for 30+ years.

You start somewhere and see the results.

Why are you so afraid?

.

davidmash · #19 08-20-2011, 12:25 AM

If you work with data and think that 10 people from two groups representing a total of about 200 million is going to get you e good sample you go right ahead. I'm just very glad you do not work for me. I'm shaking in my boots.

Botnst · #20 08-20-2011, 12:29 AM

A small fraction can indeed represent a huge number. For example, a thousand or so respondents is all that are needed for a political or commercial product poll. It's all in the sampling scheme and the underlying assumptions.

However, I agree with you that no single site (and probably not the sum of all Internet sites) is sufficient to represent the remainder of the planet.

davidmash · #21 08-20-2011, 12:35 AM

Yes, 1,000 is a typical sample size with a margin of error of about 2%, but if you look at sampling in general the small the size the larger the error. Just doing a brief web search the smallest sample size I saw on a national lever was 50 and the margin of error was 14%. At 100 it was 7%. A sample size of 10 would probably give you a margin of error of 40% to 50% which in my opinion is useless.

sjh · #22 08-20-2011, 12:45 AM

Quote:

Originally Posted by davidmash

Yes, 1,000 is a typical sample size with a margin of error of about 2%, but if you look at sampling in general the small the size the larger the error. Just doing a brief web search the smallest sample size I saw on a national lever was 50 and the margin of error was 14%. At 100 it was 7%. A sample size of 10 would probably give you a margin of error of 40% to 50% which in my opinion is useless.

If random samples for a given population, etc.

Representative sampling, targeted sampling, etc.

Further if the goal is qualitative and not quantitative large margins of error may be acceptable.

usually you take a look at the data and see if it warrants further examination.

I know what I'm talking about. I'm pretty darn good at a couple of things.

No brag, just fact.

.

el presidente · #23 08-20-2011, 08:19 AM

Subscribing....

chilcutt · #24 08-20-2011, 08:49 AM

Who am I ???

kerry · #25 08-20-2011, 08:52 AM

What difference does it make since we're all irrational anyways? At least I think I heard that argument on here recently. No, I think it was we are all non-rational anyways.

I have not noticed anyone on the left argue on this board, "Book XXX says YYY is true and since the book is inerrant there's no debating it."

MS Fowler · #26 08-20-2011, 10:36 AM

Quote:

Originally Posted by elchivito

I do not resent being placed on the left for this particular circle jerk, but wish to state that it's inaccurate. I have as many Libertarian leanings as I do Democratic and in fact generally consider myself to be a true old school conservative. Goldwater style. You remember? Conservatives used to believe that what happened in a person's home and family was none of the government's goddam##d business. That's before their brains were taken over by the Project For A New American Century. That's me, conservative thru and thru. So, if you insist on putting me on the left, be my guest. Just be aware that your study is invalid. Anyhow, let's see the list!

Interesting. I would classify myself as a Classic Liberal--meaning I put individual rights above corporate rights. I would whole-heatedly agree that "what happened in a person's home and family was none of the government's goddam##d business"

davidmash · #27 08-20-2011, 10:50 AM

I consider my self a fiscal conservative of sorts and a social liberal of sorts. I do believe in a temporary safety net, not big on wasteful spending and I believe what consenting people do is not of my business so long as it does not have a direct affect on me (not liking it does not count as an affect). I do believe that government has a valid roll in our lives such as environmental protection, a roll in health care (regulation and cost management).

So go stick that in your survey and tell me where I fit.

tbomachines · #28 08-20-2011, 11:03 AM

Quote:

Originally Posted by sjh

Discussion in the public forum.

You start somewhere.

Since the preferred mode of many is shrillness I offer an alternative.

You surely don't believe you are the only academic; this site has a dozen or more.

Adjust sample as one wants.

I suspect even with a wide latitude of settings the results will be consistent.

This allows the discussion to move from rhetorical to actuarial.

In real-world scenarios that is considered progress.

Dismissing with a few sentences, of course is an alternative.

.

I know I'm not the only one in here, I wasn't implying that and we've been over this before. Anyways to make your sAmple work you would need to know where each of these 20 people objectively land in the greater spectrum of political partisanism. Until you determine that, you have no way of verifying that they are even close to representative, nor do you have the objectivity needed to determine whether they are even "political" and to what degree. Just saying someone exhibits "left" or "right" characteristics from your point of view means nothing to me unless it has supporting data. If you want to go about this scientifically as it seems from your firs post I think you are probably not going to get too far. Anecdotally I think you have a case.

sjh · #29 08-20-2011, 07:19 PM

Quote:

Originally Posted by davidmash

I consider my self a fiscal conservative of sorts and a social liberal of sorts. I do believe in a temporary safety net, not big on wasteful spending and I believe what consenting people do is not of my business so long as it does not have a direct affect on me (not liking it does not count as an affect). I do believe that government has a valid roll in our lives such as environmental protection, a roll in health care (regulation and cost management).

So go stick that in your survey and tell me where I fit.

There are numerous tools that are readily available to quantify these discussions.

Examining if there are differences between rationales and personal perspectives is a reasonable question, even if it engenders snide responses.

Here's a study correlating personality types and their preferences for various font styles (garamond, Times Roman, etc...)

Fonts: why it takes all types

My question about socio-political values and rationales for argument is a valid topic of consideration.

,

Jorn · #30 08-20-2011, 07:24 PM

I consider my self a right wing nut.