I spent a year as a fact-checker at USA Today before embarking on a freelance career. The experience of hunting down statistics gave me a statistic of my own: the majority of oft-cited stats are completely inaccurate. We do not know that telecommuters are 40% more productive than office-based workers. We do not know that 80% of 2-income couples have hired a cleaning service. We do not know that only 5% of Americans have a valid passport. Nor do we know that 95% of small businesses fail. Sometimes authors would make the stats sound more accurate by saying "According to the Census Bureau..." or some such, but when asked for back-up, it would always be a link to another news article or someone's website, never a primary source. And, indeed, when I called the primary source, I'd get a lot of loud sighing. In the internet age, incorrect stats get bandied about a lot.
Which leads me to the topic for today's post: the statistic that 20% of high school dropouts are gifted. This stat gets used in a variety of formats, which even at first blush makes it seem problematic. Some people twist it to say that 20% of gifted students drop out. Others say "up to 20% of high school dropouts test in the gifted range" (I think I've been guilty of this one). There are old studies that mention something around this number. The 1972 Marland Report to Congress seems to have used an 18% figure. In 1973, E. Nyquist presented a paper to the National Conference on Gifted saying 19% of New York high school dropouts were gifted. E. Robertson's 1991 article in Equity and Excellence on "Neglected Dropouts: The Gifted and Talented" said 18-25% of GT students drop out. From there things get more slippery. Some people cite one Solorzano 1983 stat saying the figure is up to 18%, but this turns out to be an article in US News & World Report, not an original study.
Of course, all of this hinges on there being a clear definition of giftedness. There isn't one. Some school districts in the past used a clear 130+ IQ definition, but this method has fallen out of favor in terms of more comprehensive assessments, achievement tests and the like. There isn't even a completely uniform definition of a dropout. Some people go back to school later in life -- or at least plan to, and hence wouldn't identify themselves as dropouts. Just on an extremely long sabbatical.
All that being said, I've spent the past few hours reading a late 2002 report from the National Research Center on the Gifted and Talented about "Giftedness and High School Dropouts" that has a bit more rigor to it. The report was mainly designed to find out why gifted kids drop out. The answer is that they drop out for the same reasons other kids do. They don't like school, they're failing school, they're pregnant, they want to get a job, their parents didn't finish school, their home life doesn't particularly support learning, etc. The report found that "at risk" gifted kids -- those of lower socioeconomic status -- are more likely to drop out, as are gifted minority kids. These are unfortunate findings but not particularly surprising.
However, to establish these things, they had to find gifted dropouts. And to find them, you have to have some sense of how many of them there are. The study's authors used data from the National Education Longitudinal Study begun in 1988. This tracked 8th graders through the next few years and is the most comprehensive data set available on who stayed in school and who did not (obviously not all dropouts could be found, but many could). The authors had to use some screen for giftedness, and so they chose a relatively broad one. Students were counted as gifted if they participated in their school gifted program, or took more than 3 advanced, enriched or accelerated classes. Obviously, this is not perfect. My first high school's advanced classes would hardly have required a genius IQ to do all right in, but this at least gave the authors some screen for kids being relatively bright.
The numbers? In the first part of the study, "Among 1285 students who completed the Second Follow-Up Dropout Questionnaire, 334 were identified as gifted." That gives us a rate of giftedness among dropouts of 26%. That's even higher than the oft-cited 20%!
But, of course, it's not quite that simple. We need some context here. What is the giftedness rate among the overall population using this definition? For the second part of the study, the authors used the full cohort of 8th graders who completed all the rounds of the longitudinal survey. Of 12,625 students, 3520 met the giftedness definition, or 28%.
In other words, what this survey reveals is that students identified as gifted (by a broad definition) and other students drop out at the same rate -- and indeed, the authors did find this explicitly as well. 5% of both the 12,625 total student sample and the 3520 gifted student sample dropped out.
So what does that mean? My personal definition of giftedness is not nearly so broad as to encompass a full quarter of students. I don't think most other people believe the definition should be so broad, either. When people say "up to 20% of high school dropouts test in the gifted range" we generally use that stat to imply that the gifted are over-represented in the dropout population. We don't actually know that. I would love to see a study of students with 150+ IQs, looking at their dropout rates. The other way of getting at this -- studying all dropouts, and seeing how many have 150+ IQs -- would probably reveal the obvious. It's a very small percentage, precisely because so few people have IQs that test so high.
So if I'm reading this all correctly, to get a rate of giftedness among dropouts of 20%, you simply have to set the definition of giftedness among the total student population at 20%. If you set the definition of giftedness at 10% among the general population, you'd probably get a rate of 10% among dropouts too. This may not go all the way to 1%, or .005%, though it might. Or there might be more of a bell-curve, with top 1%-ers being less likely to drop out, but the misunderstood, lone 200+ IQ kid in the 0.001% being more likely to do so. I just don't know. But there's no reason to believe that gifted students are over-represented among dropouts.
However, it is interesting to note that being relatively bright (for instance, in the top quartile of seeming academic potential) does not protect students from dropping out. The 20% stat is usually cited to counter the belief that the gifted are all academic superstars. Many are, but some aren't, and it's important to keep in mind when looking at the numbers that many people drop out who are quite capable of not only completing high school but going on to college and even more. This is a huge waste of human potential, whatever the numbers happen to be.