Dr. Diane Ravitch attacked Campbell Brown on her blog last week. That’s old news: she’s had it in for Brown for some time. What’s new is that in this post Ravitch, who is idolized by her enormous fan club for her educational perspicacity, makes an amateur’s mistake.

Ravitch’s context is a recent video in which Brown says (accurately) that “two out of three eighth graders in this country cannot read or do math at grade level. We are not preparing our kids for what the future holds.” Here’s Ravitch’s response:

[Brown] starts by saying that 2/3 of American students in eighth grade are “below grade level” in reading and math. Apparently she refers to the National Assessment of Education [sic] Progress, the only national assessment of student skills. She confuses NAEP proficiency, a specific achievement level, with grade level.

To begin with, “grade level” is a median. Fifty percent are always above grade level, and fifty percent are always below.

Now, full disclosure: I was an English major and statistical terminology is not my strong suit. But even a lay reader like me knows that Ravitch is wrong. Grade level is never a median. It’s what teachers and other educational experts conclude is the developmentally-appropriate achievement level of a student in a particular grade. That’s what the Common Core (or whatever your state is calling it these days) represents: a list of standards-based, criterion-referenced goals that students should learn by the end of each grade in order to be ready for the next one. If the world stood still on its axis, then these goals would be the same as those 100 years ago. But the world moves and, therefore, the goalposts have changed.

In defense of her distinction between median and grade-level, Ravitch cites NAEP, certainly an authoritative source. Let’s look:

NAEP reading achievement-level descriptions present expectations of student performance in relation to a range of text types and text difficulty and in response to a variety of assessment questions intended to elicit different cognitive processes and reading behaviors. The specific processes and reading behaviors mentioned in the achievement-level descriptions are illustrative of those judged as central to students’ successful comprehension of texts. These processes and reading behaviors involve different and increasing cognitive demands from one grade and performance level to the next as they are applied within more challenging contexts and with more complex information. While similar reading behaviors are included at the different performance levels and grades, it should be understood that these skills are being described in relation to texts and assessment questions of varying difficulty.

Here’s the crux of Ravitch’s confusion. When she uses the word “median,” she means norm-referenced, i.e., how kids perform right now against the general population. In this case, indeed, half would be above and half would be below. But Brown, like NAEP, is talking about a criterion-based reference, assessing student performance based on what all these experts believe that kids should know in order to be successful in life after high school.

Here’s an example. Let’s say that a chemistry teacher is starting a unit on the structure of atoms. The teacher gives her lessons and then gives the students a test on what they learned. Is “grade level” what they happen to know after the instruction, regardless of what they learned,  or is “grade level” what they were supposed to have learned?

Answer: it’s the latter.

Ravitch must know this, right? I can see three reasons why she would make such an amateurish error:

1. She had a bad day and wrote carelessly. If any of her many readers noticed, she decided it wasn’t worth correcting.
2. She, a celebrated education expert, doesn’t know the difference between median and grade level.
3. She is deliberately misusing the two terms in order to ding Brown and promote her agenda that our schools are just fine and kids are learning what they need to know.

I can understand the first. We all have bad days. In this case, she owes it to her acolytes to issue a correction.

I don’t believe the second.

I don’t want to believe the third, but maybe I need to get more cynical.

Whatever the answer, Ravitch provides an  emblem of the American education “wars.” If we assess our traditional school system in the way that Ravitch suggests, then we can really teach anything and call it a day and cast aside meaningful expectations, core content, and effective instruction. But if we assess our traditional school system by meaningful criteria — not on what kids happen to know but what they should know — then even Ravitch knows — must know — that we come up short.