Are DC/MD/VA kids smarter than kids in other parts of the US?

[Anonymous]

Some of the tests are in fact IQ tests. WPPSI is . Yes, there is a higher percentage of smart parent and smart children. Education does not perfectly track intelligence, but it is not without correlation. Agree, parts of Boston, Silicon Valley, etc. would have similar ratios. We were told by a developmental pediatrician who also tested my child, that a 99% plus IQ related test score puts you in the top 10% here. I believe it.

[Anonymous]

OP, may I just say that I think you're fab? Your data-driven approach and no-nonsense style is much appreciated. Please let us know how the crunching turns out.

I also totally agree with the poster who wrote
"You could use this as a companion to the SES data point because there are a lot of $$ families in the NYC and LA areas especially that have high "socio" and "economic" status without having top levels of education. My cousins the Wall St. traders come to mind. "

[Anonymous]

I appreciate the short answers. If I decide to crunch the data, maybe we'll find some evidence for your views.

__________________________________________________________________________________________________________________

What data do you propose to crunch? It's hard to make chicken salad out of chicken poop. Garbage in, garbage crunched, garbage out.
I suspect you're thinking about pursuing metrics like SAT, PSAT etc. It would be very difficult to control for confounders (socio-economic drivers) that fuel prepping and anonymous sittng for exams multiple times before deciding on which try counts in the books.

[Anonymous]

OP here. Thanks to all for the useful input and kind words. I looked more closely at the NAEP data to see what I can pull from there. There are several years worth of data. Also, the data is divided into basic categories (reading, math, science, civics, etc). So far, these look like the most promising ways to slice the data:

(1) Focus on reading and math only. Those are two core subjects that everyone learns, so I think it makes sense to limit my survey to those two areas.
(2) Focus on 4th and/or 8th grade. There is data for 12th grade too, but my instinct is that lots of extraneous noise might affect 12th grade scores. Some kids will start dropping out of school, while others might be headed to college. Many kids won't be as focused on tests and schoolwork.
(3) Focus on the average 90th percentile scores. This would mean comparing the average top-10% kids in each location.
(4) Focus on the average scores of kids scoring in the "advanced" level. This seems to be about 3-7% of kids in most states/cities, so it actually is a higher-scoring group than the 90th percentile.
(5) If I get really curious, I think there is a way to separate scores within each state by population (city/suburb/town/rural). I think this might be pretty interesting, but it would be a real PITA because that's lots more data to stuff into a spreadsheet.
(6) Averaging data. Once I put all the scores into a spreadsheet, I think the easiest way to compare different states/cities will be just to average all the scores for each state/city into one raw number.

What do others think? Is there a better way to approach the data? All constructive input appreciated.

[Anonymous]

OP,
This sounds like a serious project. Are you analyzing this out of curiosity? Or for a presentation or an article? Whatever the reasons, please share you findings!

[Anonymous]

In my opinion and travels the kids from West Virginia are the smartest.

[Anonymous]

And, why does it matter? Do You just want to be able to brag or why do you care? It's a pretentious question.

[Anonymous]

I'm not from West Virginia, but from Missouri you dummie!

[Anonymous]

I don't think examining intelligence clusters is pretentious. D.C. children are sophisticated. We are reminded of this when we visit relatives in other parts of the country or have friends visit from other parts of the country. The same is true for New York, LA, Seattle, San Francisco, and other areas; D.C. hasn't cornered the market on it.

[Anonymous]

OP, you're great! I can't help with the choice of data (although I could offer suggestions about CPS and SIPP variables, but they don't cover what you're looking at). Normally I would try to control for socio-economic status, as someone suggested, but I doubt you'll find a dataset that includes this too. In any case it may be a bit beside the point -- you're not trying to prove WHY DC kids are above average, as I understand it, just that they ARE above average.

[Anonymous]

By what scientific measure?

[Anonymous]

OP here. I think some people might be confused about what I'm trying to do. To be clear, I don't know (and ultimately I don't really care) whether DC/MD/VA kids are smarter or dumber than kids in other states. I just have seen several other people on DCUM claim (without any proof) that DC/MD/VA kids score much higher on tests than kids in other places, and I want to figure out if there is any support for that claim.

Below are preliminary results from my crunching. The list below represents the average scores of the 90th percentile for each state and several cities (in other words, the top 10% of students). This is the average of the 90th percentile scores for 4th & 8th grade in 2007 and 2003. As a point of reference, the total average (median?) scores for DC/MD/VA are 222/251/255. I marked in bold and italics some interesting items. MD and VA are clearly near the top of the list, which would tend to support the claim that this area generates lots of top students. But neither of the states is too much ahead of (or behind) most other states, which suggests that this area is not too much different from most other places. Also, DC itself is at the very bottom.

Ultimately, these numbers seem inconclusive to me, largely because the statewide area they cover is too big. I think the NAEP data can be subdivided further to separate urban/suburb/town/rural within each state. That might be more useful. Maybe I'll try to do that later. In case anyone else is interested in exploring, the NAEP data is pretty interesting, and there are several pre-made tables/charts on lots of topics: http://nces.ed.gov/quicktables/index.asp .

Massachusetts 300
Connecticut 297
Minnesota 297
New Jersey 297
Vermont 296
Maryland 295
New Hampshire 295
Colorado 294
Kansas 294
Virginia 294
Washington 294
Austin 294
Charlotte 294
New York 293
Pennsylvania 293
Wisconsin 293
Illinois 292
Indiana 292
Iowa 292
Maine 292
Montana 292
Nebraska 292
North Carolina 292
Ohio 292
Oregon 292
North Dakota 291
South Dakota 291
Wyoming 291
Alaska 290
Delaware 290
Idaho 290
Michigan 290
Missouri 290
Utah 290
Florida 289
Texas 289
Kentucky 288
Rhode Island 288
South Carolina 288
Arkansas 286
Georgia 286
Arizona 285
California 285
Tennessee 285
Oklahoma 284
New York City 284
San Diego 284
Hawaii 283
Nevada 283
West Virginia 283
Boston 282
Alabama 281
Louisiana 280
New Mexico 279
Houston 278
Mississippi 277
Atlanta 272
Chicago 272
Los Angeles 270
District of Columbia 267
Cleveland 265

Here's a link to the full data: http://spreadsheets.google.com/ccc?key=0AqmbmeXDw6gxdHhldXlDeUdCNGc3d2NfYmg3ZGhZaHc&hl=en

[Anonymous]

OP again. I figured out how to subdivide the data. Within each state, the 90th percentile scores can be subdivided by population: city/suburb/town/rural. Below are summary results for each area, showing the 90th percentile scores. Un-summarized data is at the link I provided earlier. Within each location, I averaged the results for 4th grade reading from 2007, 4th grade math from 2007, and 4th grade math from 2009.

Here are the top 20 locations, where the highest 90th percentile scores (listed 1-20):
286 Vermont Suburb
284 Vermont City
283 Massachusetts Rural
282 New Jersey Rural
282 Massachusetts Suburb
282 Connecticut Rural
281 Vermont Rural
281 Maryland Town (rank 8)
281 Maine Suburb
280 Kansas Suburb
280 Pennsylvania Suburb
280 Minnesota City
279 Alaska Town
279 Connecticut Suburb
279 Vermont Town
278 New Hampshire Suburb
278 Wisconsin Suburb
278 New Jersey Suburb
278 Maryland Rural (rank 19)
278 Virginia Suburb (rank 20)

Here are 90th percentile scores from other locations people here might find interesting (ranking out of 206 total in parentheses):
277 National Private Suburb (rank 24)
276 National Private City (rank 31)
275 National Suburb (rank 37)
275 Maryland Suburb (rank 39)
274 National Private Rural (rank 41)
274 Maryland City (rank 50)
271 Virginia City (rank 85)
270 National Private Town (rank 96)
270 National Rural (rank 106)
269 National City (rank 122)
268 Virginia Town (rank 126)
268 National Town (rank 140)
267 Virginia Rural (rank 146)

Here are the bottom-three scoring places (ranks 204-206):
254 Mississippi Town
254 District of Columbia City (rank 205)
251 Mississippi City

What do I draw from all this? It seems clear the DC/MD/VA metro area is not producing any disproportionate number of 90th percentile scorers. I suspect most of this region's students would be classified DC city, MD suburb, or VA suburb. Perhaps MD town or VA town as well. No huge number of 90th percentile scorers are coming from those regions. Most seem to be in the top two-thirds, but that's nothing spectacular. Indeed, the top-20 scoring areas seem to be largely dominated by lightly-populated Northeastern and Midwestern places. Cities do not fare well.

Based on this data, my best guess is that people are engaged in a little wishful thinking when they claim DC/MD/VA students are clearly smarter than most other students across the country.