Kevin's Watch

I take it no Factor Analysis will be used on such a small sample?

Heh, I know one female in the 40-50 age group who answered...

aliantha wrote:Hey! I'm not a cynic!

I'm a realist. There's a difference!

Didn't say you were a cynic, aliantha. In fact, you were among the original Watch members who inspired me to join - back when KW could still be called a "family." Now, it's...I don't know what it is.

*shrug* It's still my family. It's changed and changed and changed, even in the short 4.5 years I've been here. But it's remained one of the bigger parts of my life. A large percentage of the best people I've ever known are here, and I've even had the great pleasure of meeting some of them in person.

Loremaster wrote:I take it no Factor Analysis will be used on such a small sample?

I have no idea what that is, but I guess the small number rules it out? I'm sure Raen will be here today or tomorrow to answer this one. heh

Matrixman wrote:Now, it's...I don't know what it is.

A bigger family.

That's all.

drew wrote:

Matrixman wrote:Now, it's...I don't know what it is.

A bigger family.

That's all.

www.pewinternet.org/pdfs/PIP_Communities_Report.pdf


Here is an official study done in 2001. I only have the attention span to read part of it but it is interesting. For those with absolute boredom on their hands

I'm too tired to be bored at the moment!

Yikes! That's a long paper!

Fist and Faith wrote:

Loremaster wrote:I take it no Factor Analysis will be used on such a small sample?

I have no idea what that is, but I guess the small number rules it out? I'm sure Raen will be here today or tomorrow to answer this one. heh

Small sample equates to an inflated variance for multiple correlations. It was a factor that plagued my thesis study.

Yeah, I understand the part about small samples making for unreliable results. But I never heard of "Factor Analysis." Is it a way to try to make up for small samples?

Fist and Faith wrote:Yeah, I understand the part about small samples making for unreliable results. But I never heard of "Factor Analysis." Is it a way to try to make up for small samples?

A factor analysis is a statistical method used to 'tease' apart correlations (patterns) in a matrix of intercorrelations. However, small samples tend to produce spurious, or inflated, results.

I don't understand "'tease' apart." An attempt to prove that apparent correlations are not... "real" correlations?

Fist and Faith wrote:I don't understand "'tease' apart." An attempt to prove that apparent correlations are not... "real" correlations?

To find connections - co-relating variables. Say, a relationship between muscle strength and body size.

So a Factor Analysis looks for correlations? OK, we're having a needless conversation. I thought you were looking for something else.

Fist and Faith wrote:So a Factor Analysis looks for correlations? OK, we're having a needless conversation. I thought you were looking for something else.

It looks for patterns in correlations - factors.

Matrixman wrote: Didn't say you were a cynic, aliantha. In fact, you were among the original Watch members who inspired me to join - back when KW could still be called a "family."

Really? Gawrsh, MM.... (And I knew you weren't calling me a cynic. I just saw an opportunity to use one of my favorite one-liners....)

Matrixman wrote: Now, it's...I don't know what it is.

It's a dysfunctional family, of course. And yes, we do put the "fun" in dysfunctional.

aliantha wrote: It's a dysfunctional family, of course. And yes, we do put the "fun" in dysfunctional.

That puts things in perspective. Thanks.