[personal profile] philobiblius
One of the questions that arises along the way is "Are gifted children at risk?" Inevitably, the question brings the reply "At risk of what?"

Dropping out, depression, drugs, delinquency, and death (self-inflicted) are the answers. 4 of these five are pretty commonly discussed within the gifted lit - and often researchers seek to support or refute them. There are lots of stories and fewer statistics - and what stats there are may be misquoted, misremembered, or misremembered.

For the moment, I am going to focus on delinquency. I've been reading one of the studies that set out to disprove the notion that violent adolescents are any likelier to be gifted than the general population: High intelligence and severe delinquency: Evidence disputing the connection, by Dewey G. Cornell, in Roeper Review, May 92, Vol. 14, Issue 4.

Dr. Cornell had 157 violent offenders to examine and he did a pretty thorough job of illustrating his general point. Of the 157, "only 2 subjects obtaining scores greater than 130, and 2 more scoring greater than 120" on their full scale IQs (WISC-R or WAIS-R). He took it further, correctly observing that prison populations are known for higher performance scores than verbal.

"There were 13 subjects with Performance IQ's of at least 120. This included two subjects with IQ's greater than 140 and two more with IQ's greater than 130. In contrast, there were only 3 subjects with Verbal IQ's of 120 or higher, and all 3 had equivalent or higher Performance IQ's."

13 out of 157 is not overrepresented for 120+ (9% is the expected percentage.)

He talked a bit about race, and looks at the fact that the 'minority' members of the 13 above 120 performance IQ group were only 31% (4 of 13) vs. being 75% of the below 110 population. From there, he continued to explore his 2nd question, "Do highly intelligent delinquents differ from other delinquents in their social background and prior adjustment?"
*******

And that is where I think Dr. Cornell made his mistake.

The white population of the total 157 group was 44, or 28% of the whole. The number of whites who scored 120 and above on the Performance Scale was 9, or more than 20% of the white population, when 9% would have been expected.

Cornell wrote, in conclusion: "The results of this study provide evidence that high intelligence is not associated with severe delinquency. In fact, the majority of delinquents are of below average intelligence, and only a few delinquents obtained scores above the high average range. While it is possible to identify delinquents with high intelligence, it is not reasonable to infer a connection between delinquency and high intelligence."

I think he missed a vital segment of his population.
********

This is hardly conclusive to prove risk, let alone to be as definitive in the opposite direction from Cornell. But it does at least raise an unanswered question: Might there be a greater risk for gifted (high performance scale) white adolescents to become seriously delinquent than for the norm?

(originally written Aug. 27th, 2010)

Postscript: In addition, there are factors to be considered, including size and location of population. There are also language issues deeply embedded in the prison population, perhaps tied to the number of dyslexics (or, as Ken Seeley would note, visual-spatial individuals).

Date: 2013-07-30 05:12 pm (UTC)
staranise: A star anise floating in a cup of mint tea (Default)
From: [personal profile] staranise
Yes, I definitely want to see more analysis of race here. Oh, of everything. These are violent offenders, not the prison population at large; the Wechsler tests already tend to be wonky around race; does the ratio of high-IQ people of various races change if you move between offender types? Does our society place pressures on races differently so that high-IQ whites are more prone to certain activities than high-IQ black people?

Can we pretty pretty please have another large-scale longitudinal study of gifted children?

Profile

philobiblius

February 2017

S M T W T F S
   1234
567891011
12131415161718
19202122232425
262728    

Most Popular Tags

Page Summary

Style Credit

Expand Cut Tags

No cut tags
Page generated Jun. 25th, 2017 05:28 pm
Powered by Dreamwidth Studios