Virginia Sets the Standard for Dropout Reporting

Here is a thought experiment: Imagine what life would be like if all rulers were relative. One person measuring a table might say it was 4 feet long. Another, using a different ruler, might say it was 11 feet long. Who's right? Both, or neither? It would be impossible to say. Now imagine two other people measuring a different table across the country with two more unique rulers. Is the second table bigger than the first, the same size, or smaller? Again: impossible to say.

Something like that problem has been going on in public education. States and localities have long measured their own dropout rates in different ways. That hugely complicated the business of comparing performance across school divisions, demographic groups, and retention programs.

Even worse, some school districts effectively have kept two sets of books -- one that went to Washington for reporting purposes, and another that reflected something closer to reality. In a particularly notorious case, Jackson, Miss., was reporting a graduation rate of 99 percent even though half of its entering freshmen vanished during the following four years.

The problem was so vexing that in 2005 the National Governors Association entered into a compact to redress the matter. It evidently failed to do so -- because two years ago Virginia Rep. Bobby Scott introduced the Every Student Counts Act, which also sought to improve graduation-rate reporting. And about this time last year, the Bush administration's education secretary, Margaret Spellings, announced similar plans to take administrative steps that would impose a uniform nationwide standard.

BUT JUST as Virginia's Standards of Learning preceded the No Child Left Behind Act, Virginia's Department of Education has been out front on graduation rates. The other day the department released figures from a program that tracks dropout rates by assigning every student a unique identifier. The method allows the commonwealth to know whether a student who enrolls in ninth grade receives a sheepskin four years later. That provides a far more accurate snapshot than, say, counting the number of entering seniors in a school district and then counting the number who graduate nine months later.

Measuring the dropout rate on the basis of a single year can mask the scope of the problem in the following way. Suppose a school enrolls 100 freshmen. By the end of the academic year, 10 students have dropped out and only 90 move up to the sophomore class. Over the course of the sophomore year, 10 more students fall away. Ten more do so in their junior year, and 10 more in their senior year. If you average the individual dropout rates for each of the four years, you get a dropout rate of only 11.8 percent. But over the course of four years, 40 students dropped out -- for a cumulative rate of 40 percent.

And if five students move into the school district each year, that could mask the apparent size of the dropout rate as well: A freshman class starts with 100 students, loses 10, but picks up five, ending the year with 95 students -- for (perhaps) an apparent dropout rate of only 5 percent. What's more, suppose that all 10 students who drop out are African-American (or Latino, or Asian -- whatever). Merely counting heads papers over a problem needing redress.

THE DROPOUT data released Tuesday don't deliver any bombshells. Minorities, poor students, and those with disabilities or limited English proficiency drop out more frequently. Petersburg faces huge challenges. Hanover doesn't.

On the other hand, the numbers might help to dispel some misperceptions. Richmond's dropout rate of 16.2 percent is not grotesquely higher than that of, say, Chesterfield (11.7 percent) -- although their on-time graduation rates diverge more widely (65.9 percent versus 84.7 percent). Perhaps, given the demographic challenges it faces, the Richmond school system is doing a better job than stereotype would suggest. Or perhaps Chesterfield has bigger problems than some residents realize. The gap between a district's graduation rate and dropout rate is filled by, among other things., "completers" who don't drop out but who don't earn a diploma, either.

It is axiomatic that you cannot improve what you do not measure. And a corollary might hold that inaccurate or inadequate measurements might be nearly as bad as none at all. Virginia's more accurate method of counting dropouts means, as Board of Education president Mark Emblidge said, that "educators and policymakers can now see where interventions are most urgently needed and identify high schools and school divisions that have developed best practices and strategies that others can emulate and adapt."

Implementing a suitable statistical methodology for measuring high-school dropout rates is not one of those issues likely to light up the phone lines of talk radio or cable-TV news. But apparent dullness is often inversely proportional to actual importance. For those who toil in the educational vineyard, improving dropout rates is not about merely massaging statistics; it is about sparing the young from a lifetime of regret.

If we could read the secret history of our enemies, we should find in each man's life sorrow and suffering enough to disarm all hostility. --Henry Wadsworth Longfellow.

Contact A. Barton Hinkle at (804) 649-6627 or .