EXTRA CREDIT PROBLEM. For research I'm doing, I'd like to explain the differences across U.S. states in the fraction of K-12 students classified as "gifted and talented".

I assume initially that in large groups of students, the fraction who are gifted and talented should be about the same. I say "about" because external conditions--poverty, for one--may prevent some students from reaching their potentials and also because high IQ workers are probably unequally distributed across states and so, therefore, would their children.

I further assume that most of the variation seen in the table linked to above is therefore due to differences in how "gifted and talented" students are identified which depends, in turn, on the educational systems of the states. Some support for this is that the simple correlation between the fraction of adults who hold graduate or professional degrees and the percentage of gifted and talented students is -.08; the fraction of adults who have bachelor's degrees, -.06; and median household income, -.03.

Some potential causal factors I want to explore are teacher unionization; political attitudes; economic and social inequality; and availability of private schools (which in the long run also would need to be explained).

As a quick-and-dirty test of the first two factors, I computed the correlation with the fraction of the electorate voting for Bush in 2004. I get .135; .195 without Maryland.

Suggestions welcome.

Comments

[Joshua Hall]

You might want to look for whether a state has a requirement that kids be tested. I know that Ohio was considering such a rule at one point, although I don't know if it was implemented.

[Jake]

In business, there is correlation of high intelligence with high income. It would be interested to see whether high income states have more gifted students.

But I think you are working with data that is so faulty that no meaningful analysis can be done between states.

Look at Massachusetts with 1.2% and Vermont with 1.1% gifted. Normally, I would say that the two most left-wing states in the nation are also the dumbest. A better explanation is that their left-wing educational establishment does not measure gifted students as it would hurt the self-esteem of the non-gifted students.

A better idea is to pick a state (or states) that have a consistent statewide program for measuring gifted students. Then do your analysis by school district or county.

[SteveSC]

A couple of interesting tidbits from eyeballing the data:

1) Overall, selection as gifted is biased towards girls, with 6.7% of girls in the US labeled gifted and only 6.0% of boys. This is likely a selection bias due to testing, and most likely it is because girls are more verbal and the testing overweights verbal ability.

2) The states with the highest percentage of gifted overall have the greatest bias towards girls. In fact, a quick Excel calculation shows a 77% correlation of higher percent gifted with higher girl minus boy percentages.

I would hypothesize that states with a more intensive screening program are more likely to 'identify' gifted, and are likely using paper/pencil testing rather than more extensive professional testing.

3) If you rank the states by the percentage of gifted, the top 10 are Oklahoma (220% of US percentage), Maryland (189%), Kentucky (183%), Nebraska (178%), Arizona (165%), Virginia (163%), North Carolina (159%), Wisconsin (158%), South Carolina (156%), and Texas (142%). The bottom 10 are Vermont (18%), Massachusetts (19%), New Hampshire (28%), Wyoming (29%), West Virginia (34%), Rhode Island (34%), North Dakota (38%), Utah (46%), Tennessee (47%), and Kansas (52%). I guess there is some red state v. blue state diffference there, but it doesn't seem to be exclusive.

My thought in a mess like this is to follow the money. Which states financially reward finding 'gifted' students? This may be relatively independent of the factors you are looking at.

[SteveSC]

A couple of interesting tidbits from eyeballing the data:

1) Overall, selection as gifted is biased towards girls, with 6.7% of girls in the US labeled gifted and only 6.0% of boys. This is likely a selection bias due to testing, and most likely it is because girls are more verbal and the testing overweights verbal ability.

2) The states with the highest percentage of gifted overall have the greatest bias towards girls. In fact, a quick Excel calculation shows a 77% correlation of higher percent gifted with higher girl minus boy percentages.

I would hypothesize that states with a more intensive screening program are more likely to 'identify' gifted, and are likely using paper/pencil testing rather than more extensive professional testing.

3) If you rank the states by the percentage of gifted, the top 10 are Oklahoma (220% of US percentage), Maryland (189%), Kentucky (183%), Nebraska (178%), Arizona (165%), Virginia (163%), North Carolina (159%), Wisconsin (158%), South Carolina (156%), and Texas (142%). The bottom 10 are Vermont (18%), Massachusetts (19%), New Hampshire (28%), Wyoming (29%), West Virginia (34%), Rhode Island (34%), North Dakota (38%), Utah (46%), Tennessee (47%), and Kansas (52%). I guess there is some red state v. blue state diffference there, but it doesn't seem to be exclusive.

My thought in a mess like this is to follow the money. Which states financially reward finding 'gifted' students? This may be relatively independent of the factors you are looking at.

[John Chilton]

The most recent "best of the web" carried a story about a blue state school district that was doing away with spelling bees as inconsistent with the policy of no child left behind. Seems like a case of foolish consistency and hobglobbins of the small mind.

There's increasing evidence that some middle and upper middle income parents try a strategy at the other end -- to get their child classified as learning disabled so they will be allowed more time on tests and thus have a better chance of getting into a good university. I guess universities look at test scores and not whether you're disabled???

I wonder if looking at cases of mergers of school districts would be helpful in testing an idea along the lines of economics and social inequality with a twist. My notion is that when more diverse groups are thrown into the same school you should see the number of students labeled as gifted rise as a way to achieve segregation (whether because of classism or because of the belief that less learning is accomplished in more diverse classrooms).

[Mark Shroder]

As a Marylander, I can assure you that Lake Wobegon is in my state, not Minnesota.
Any student who isn't rated "gifted and talented" has parents who won't lobby for her.