There has been a great deal of discussion lately about the latest school test scores.

Some schools, as you’ll see elsewhere in this week’s paper, performed well on some tests, but poorly on others. Other schools did well on every test. Still others performed poorly no matter what the measure.

Such scattershot results make it hard to imagine that testing — as its now constructed — is worth a lot.

The feds have their No Child Left Behind. That’s a noble idea, but it doesn’t take into account the notion that some children have issues far too complex to expect them to perform well on academic tests.

Those issues range from the namby-pamby bleeding-heart-liberal idea that some children have poor home lives and aren’t getting the support they need at home to do well in school to students with profound learning disabilities. There are also students who have found their way to North Carolina from points south of the U.S. border who must learn the English language before they can be expected to succeed in school.

The state, meanwhile, has its eye on improvement. If a school did poorly last year, then how did they do this year? If student performance got better, the argument goes, then teaching must have gotten better. But that doesn’t take into account the notion that you’re comparing different children. Last year’s crop of fifth-graders, for instance, could have been a bunch of relative dummies, while this year’s class was full of geniuses.

And, both the state and the feds play an interesting game of switcheroo from time to time, changing the testing model so that prior years can’t be compared to the current year. That buys school systems time to retool and consider new ways to make students perform better on tests.

Ultimately, testing is the most efficient way to tell if students are learning and teachers are teaching. Injected with a little common sense, testing could be a much more effective tool than it already is. But whose model should we follow? Federal guidelines only apply to schools that receive federal funds. The U.S. government has never really been interested in dipping its toe into education. If the feds’ No Child Left Behind made some kind of exception for students with truly overwhelming obstacles, that might be a reasonable measure. States have long been responsible for education. And in multiple rankings, some states spend more per student than others, while other states pay their teachers a higher annual salary.

There are myriad ways to determine that your state’s schools are better or worse than some other state’s schools.

It can be, in a word, dizzying.

North Carolina’s testing model considers improvement. It’s a model other states have learned and copied in one form or another.

If state officials would figure out a simple way to measure improvement between this year’s eighth-graders and last year’s seventh-graders — in other words the same students — parents and taxpayers might have a better idea of just how well the schools are doing. Wherever your child’s school ranked in the most recent survey, two things are certain: nobody really knows what it means, and school starts on Monday.

So drive carefully.