Consider this story problem:

Sally has a load of 100 apples to deliver to market. So does Johnny. Sally fills her wheelbarrow, but along the way 20 apples fall out. She reports a dropout rate of 20 percent. Johnny fills his barrel with 90 apples, failing to include 10 of them. Along the way 9 apples fall out. He reports a dropout rate of 10 percent. What’s wrong with Johnny’s math?

The answer underlines the various ways dropout rates are calculated across the country. The good news is that Secretary of Education Margaret Spellings is onto this ruse and will issue a federal edict later this year that requires the same formula be used to calculate dropout rates at all public high schools.

For the purposes of accountability, states have long been required to report dropout rates, but each state was left to determine how the calculation would be done. A recent New York Times article highlighted the absurdity of creative counting.

New Mexico counts the percentage of enrolled 12th-graders, which erases all the students who dropped out before their final year. Using the formula preferred by feds, New York’s rate would be 35 percent, but its reported rate is 23 percent. Mississippi has two sets of books. The more accurate rate is 37 percent; the one reported to the feds is 13 percent.

Closer to home, districts use a variety of methods to lower their numbers. For instance, West Valley, East Valley and Central Valley school districts report dropout rates that are far below the state average, but that’s partly because they’ve moved many at-risk students into alternative programs, where only 22 percent graduate on time. Spokane Public Schools says its dropout rate is 11.5 percent. The state disagrees, saying it is 7.9 percent.

The feds’ preferred calculation reaches back to ninth grade, because that’s where the dropout problem generally begins. By excluding ninth-graders from calculations, educators miss an opportunity to discover why high school isn’t working for those kids.

While inflexibility is good for assessing dropout rates, it is not good for assessing the progress schools are making. Spellings is also addressing educators’ longstanding complaints with “average yearly progress” reports. To receive a passing grade and avoid federal penalties, schools must show progress among all classifications of students. A school with generally improving scores can still be judged failing if the score for a particular subset of kids doesn’t meet the target. Classifications include special needs, income and ethnicity.

But the point of assessing school performance is to direct the attention where it is most needed. Schools with failing scores across the board need more help than those with one subset of kids not meeting targets. Current AYP measures distort the picture and treat all failures equally.

Spellings seems to recognize this flaw by instituting a pilot project in 10 states that will allow districts to concentrate on the worst performing schools. In turn, the feds will withhold the “failing” label from schools that are generally doing well.

School reform is complicated, so there’s no shame in common-sense tweaking. Significant goals can still be achieved – as long as Sally does the counting.