Welcome to AikiWeb Aikido Information
AikiWeb: The Source for Aikido Information
AikiWeb's principal purpose is to serve the Internet community as a repository and dissemination point for aikido information.

Sections
home
aikido articles
columns

Discussions
forums
aikiblogs

Databases
dojo search
seminars
image gallery
supplies
links directory

Reviews
book reviews
video reviews
dvd reviews
equip. reviews

News
submit
archive

Miscellaneous
newsletter
rss feeds
polls
about

Follow us on



Home > AikiWeb Aikido Forums
Go Back   AikiWeb Aikido Forums > Open Discussions

Hello and thank you for visiting AikiWeb, the world's most active online Aikido community! This site is home to over 22,000 aikido practitioners from around the world and covers a wide range of aikido topics including techniques, philosophy, history, humor, beginner issues, the marketplace, and more.

If you wish to join in the discussions or use the other advanced features available, you will need to register first. Registration is absolutely free and takes only a few minutes to complete so sign up today!

Reply
 
Thread Tools
Old 10-22-2009, 07:30 PM   #76
Keith Larman
Dojo: AIA, Los Angeles, CA
Location: California
Join Date: Apr 2005
Posts: 1,604
United_States
Offline
Re: "Discrimination".

Quote:
Mike Sigman wrote: View Post
I gotta admit... that's one of the funniest posts that I've read in a while, David. So a personal anecdote is not a personal anecdote if it's told about something that happened to Keith during his working hours?
Gotta say this is the first time anyone has waved away my professional experience as lead researcher on a number of large scale studies so lightly. No, I didn't provide references, but would you like me to? Heck, you might even be able to find fragments of my past on google. Try googling my name and Psychometrics. My title was technical director. I ran large scale studies involving a number of educational institutions and companies like Fedex, GTE data services, Sprint, governmental agencies, and others. We even reworked, pilot tested, validated, etc. the ASVAB for the military. I was interviewed any number of times by print media, but most of those were before the Internet was so pervasive so I'm not sure what you will find. And most were in rather geeky publications.

I've probably still got copies of many of the studies I performed. And some of the summary data appears in information packages the company provided to clients.

And Mike, you may be surprised to find out I agree with you on Asian tending to score higher on standardized tests. That's not exactly news in the testing field. Neither is the fact that minority groups often score lower even on tests without any sort of "racial bias". Why these things happen are interesting discussion, but they tend to explode rather quickly.

You posted a bloggish entry of a guy asserting something of interest to me. I found it an interesting topic. Yes, I read it. Closely. I agree completely that so-called "soft" criterion should be looked at. Nepotism, racism, cronyism, etc. are all wrong. But the author looked at discrepancies of SAT scores. Good god, talk about a red flag for anyone in the testing field. I'll try to stick to well documented (i.e., easy to look up yourself) values for this. The SAT, according to published studies, has a correlation reported ranging from .30 to .38. That ain't all that impressive. But you may wish to consider my opinion merely an uninformed opinion. Okay, a pearson product moment correlation of .38 (we'll assume the highest validity to give them the benefit of the doubt) works out to accounting for 14.44 percent of variation of the predicted variable (end of Freshman Year Grades). To compute this yourself look up pearson product moment correlation and look for variation. Or you could take my word for it that all you have to do is square the pearson. Either way... So, over 85% of the variation is not accounted for. And the thing being predicted is end of freshman year grades. So we need to ask how stellar that is as a metric of success. Not bad, I suppose, but when you can only account for under 15% of the variation...

Next point would be how tests like these are used for selection. Tests like these tend to be used as coarse cut scores because they are simply not very good predictors in the first place. So selection criterion *other than* the test is what is used to pick among those who pass the first hurdle.

FWIW a bunch of years ago (and no, I'm not going to dig up references for you) I was asked to look at data for Caltech by a local reporter doing a story. At the time the percentage of asians in the US population (this was around 1998 if memory serves) was around 4%. The percentage of asian enrolled as students at Caltech was around 20%. At the time I compared it with reported data for the UC (University of California) system. The number there was closer to 30%.

The reporter didn't bother with his story. FWIW I just googled the demographics this afternoon and as of now, Asians outnumber whites in the UC system 40% to 38%.

Now, before you get upset at me none of this means anything outside of those schools. It is entirely possible there is ethnic discrimination of Asians at any number of other schools.

But again, I posted what I did because I was a testing geek. And I simply wanted to point out the folly of using SAT scores as some sort of indicator of a problem.

That doesn't mean the problem doesn't exist. But SAT scores aint' exactly a good way to show much.

And fwiw I used you and me as examples in a simple comparison. I was simply trying to say that two reasonable, intelligent and sincere people can differ on whether an individual is qualified. How you could take that as somehow being a prelude to an ad hominem attack is beyond me.

Anyway, I'm done on this kind of thread. Ironically I actually agree with many of the issues you raised about transparency. I even worked on trying to get that with a couple clients of mine in the education world back in the good old days. My mistake was forgetting that threads seem to need to devolve into us vs. them diatribes. And I was just trying to discuss some geeky psychometric points about correlations, test quality, predictive power, and group statistics.

Obviously the wrong place for this.

References available upon request...

Last edited by Keith Larman : 10-22-2009 at 07:36 PM.

  Reply With Quote
Old 10-22-2009, 07:35 PM   #77
Keith Larman
Dojo: AIA, Los Angeles, CA
Location: California
Join Date: Apr 2005
Posts: 1,604
United_States
Offline
Re: "Discrimination".

I should also throw out that for those who'd like to read more but don't want to read dry books on psychometric theory or industrial psych, Malcolm Gladwell's latest book, Outliers, has a chapter or two discussing different cultural approaches to education, studying, etc. He goes into some detail about the positive effects this has on Asian students. You'd probably like the book a great deal, Mike. I certainly did and find no fault in what he wrote.

Now the question I'm pondering is how anyone can post anything without it being just an anecdote... If I have to start posting references the way I used to this place is going to be dryer than the Sahara...

  Reply With Quote
Old 10-22-2009, 07:50 PM   #78
Mike Sigman
Location: Durango, CO
Join Date: Feb 2005
Posts: 4,123
United_States
Offline
Re: "Discrimination".

Quote:
Keith Larman wrote: View Post
Gotta say this is the first time anyone has waved away my professional experience
Wait a minute, Keith. Do you understand what "anecdotal evidence" is in regard to a bona fide study? If I cite a case in which I found that a Minnesotan was "below average in I.Q." I cannot extrapolate from that anecdotal case that "all Minnesotans are dumb". In other words, anecdotal tales are almost meaningless, no matter who did them. Do you understand that?
Quote:
And Mike, you may be surprised to find out I agree with you on Asian tending to score higher on standardized tests. That's not exactly news in the testing field. Neither is the fact that minority groups often score lower even on tests without any sort of "racial bias". Why these things happen are interesting discussion, but they tend to explode rather quickly.
Regardless, all studies must be part of an open discussion. What I react to is any attempt to not discuss any data because it might be "racist". My sensors go on full alert when someone calls for any legitimate data to be ignored... particularly when they try to use intimidation (real or implied) in their comments. There are some interesting *anecdotes* I can provide to show what attempting to hide factual evidence winds up doing in the long term.
Quote:
You posted a bloggish entry of a guy asserting something of interest to me. I found it an interesting topic. Yes, I read it. Closely. I agree completely that so-called "soft" criterion should be looked at. Nepotism, racism, cronyism, etc. are all wrong. But the author looked at discrepancies of SAT scores. Good god, talk about a red flag for anyone in the testing field. I'll try to stick to well documented (i.e., easy to look up yourself) values for this. The SAT, according to published studies, has a correlation reported ranging from .30 to .38. That ain't all that impressive. But you may wish to consider my opinion merely an uninformed opinion. Okay, a pearson product moment correlation of .38 (we'll assume the highest validity to give them the benefit of the doubt) works out to accounting for 14.44 percent of variation of the predicted variable (end of Freshman Year Grades). To compute this yourself look up pearson product moment correlation and look for variation. Or you could take my word for it that all you have to do is square the pearson. Either way... So, over 85% of the variation is not accounted for. And the thing being predicted is end of freshman year grades. So we need to ask how stellar that is as a metric of success. Not bad, I suppose, but when you can only account for under 15% of the variation...
Then there's this:

http://www.law.ucla.edu/sander/syste...anderfinal.pdf
Quote:
That doesn't mean the problem doesn't exist. But SAT scores aint' exactly a good way to show much.
So either testing of some sort is valid or not, right? Do you suggest that all testing is non-valid?
Quote:

And fwiw I used you and me as examples in a simple comparison. I was simply trying to say that two reasonable, intelligent and sincere people can differ on whether an individual is qualified. How you could take that as somehow being a prelude to an ad hominem attack is beyond me.
I think an argument can be made without interjecting any personal comments or perspectives, Keith. Actually and literally, that's a fairly common view, in terms of legitimate debate.
Quote:

Anyway, I'm done on this thread. Ironically I actually agree with many of the issues you raised about transparency. I even worked on trying to get that with a couple clients of mine in the education world back in the good old days. My mistake was forgetting that threads seem to need to devolve into us vs. them diatribes. And I was just trying to discuss some geeky psychometric points about correlations, test quality, predictive power, and group statistics.

Obviously the wrong place for this.

References available upon request...
Good comments, Keith... plus you sustained your argument from a number of perspectives, which is important to me. The one common factor I note on a lot of internet forums is the presence of "Gollums" (As Bilbo ran, Gollum cried out, "Thief! Thief, Baggins! We hates it, we hates it, we hates it forever!" ) who make no attempt at legitimate argument, as opposed to people who argue legitimately. There are a couple of what I would even call "Super-Gollums" on AikiWeb, BTW. Regardless, as long as someone debates legitimately, I have no problem with the dialogue. What's depressing is the number of people in martial-arts who talk about "Zen" and all that but who in reality are simply people wearing "look at me" uniforms and politics. At least you argue legitimately.

Best.

Mike
  Reply With Quote
Old 10-22-2009, 08:00 PM   #79
Keith Larman
Dojo: AIA, Los Angeles, CA
Location: California
Join Date: Apr 2005
Posts: 1,604
United_States
Offline
Re: "Discrimination".

Yes, Mike, I know what anecdotal evidence is. Thank you.

My posts were trying to just have a pleasant conversation about a topic that used to be of interest to me on a professional level. And a pretty common argument that appears is about perceived discrimination based on SAT scores. It is often used but few seem to understand how SAT's are used, what the scores mean, and how universities select students. I've said more than I needed to. And I think i've forgotten that my enthusiasm for the excessively dry field I used to be in isn't necessarily shared. No ulterior motives. No concerns. I had no emotional attachment to it. Still don't.

Carry on...

  Reply With Quote
Old 10-22-2009, 08:05 PM   #80
Mike Sigman
Location: Durango, CO
Join Date: Feb 2005
Posts: 4,123
United_States
Offline
Re: "Discrimination".

Quote:
Keith Larman wrote: View Post
My posts were trying to just have a pleasant conversation about a topic that used to be of interest to me on a professional level. And a pretty common argument that appears is about perceived discrimination based on SAT scores. It is often used but few seem to understand how SAT's are used, what the scores mean, and how universities select students. I've said more than I needed to. And I think i've forgotten that my enthusiasm for the excessively dry field I used to be in isn't necessarily shared. No ulterior motives. No concerns. I had no emotional attachment to it. Still don't.
I feel the same way, Keith. I'd argue with Satan about the correct direction to Hell. If I knew the real direction and he was unsure, I'd make a bet so I could make a few bucks.

Mike
  Reply With Quote
Old 10-23-2009, 01:41 PM   #81
Ron Tisdale
Dojo: Doshinkan dojo in Roxborough, Pa
Location: Phila. Pa
Join Date: Jun 2002
Posts: 4,615
United_States
Offline
Re: "Discrimination".

Quote:
Mike Sigman wrote: View Post
I feel the same way, Keith. I'd argue with Satan about the correct direction to Hell. If I knew the real direction and he was unsure, I'd make a bet so I could make a few bucks.

Mike
Now THAT I believe! But what's with the personal comments?? Just kidding...

Best,
Ron

Ron Tisdale
-----------------------
"The higher a monkey climbs, the more you see of his behind."
St. Bonaventure (ca. 1221-1274)
  Reply With Quote

Reply


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off


All times are GMT -6. The time now is 12:00 AM.



vBulletin Copyright © 2000-2024 Jelsoft Enterprises Limited
----------
Copyright 1997-2024 AikiWeb and its Authors, All Rights Reserved.
----------
For questions and comments about this website:
Send E-mail
plainlaid-picaresque outchasing-protistan explicantia-altarage seaford-stellionate