Data dumpster dive! Today’s aquatic event: New test scores from Maryland
Numbers too small to read? Right click on the “zoom in” control. Still confusing? Oh well, still looks good as a graphic. Here’s what you’re looking at — the gender breakdowns for the new Maryland State Assessment numbers the Washington Post has been praising for two days straight. Big gains all around, including for minority students.
(Not sure why the good-news MSA hasn’t drawn the same skepticism triggered by the equally good-news New York tests just released. It’s not easy to do anything in NYC.)
Drum roll for the Maryland results by gender: Girls in Maryland are narrowly outscoring boys in math and maintaining a distant lead in reading — exactly the pattern discovered in the two previous dumpster dives. Here are the results of the dive into New York test numbers; here are the results for the Minnesota dive.
Have the Maryland gaps narrowed over the years as overall scores rose? Not by much, as you can see in these archives that go back to 2003 (before then, the state used a different test).
Ok, there are 47 more data dumpster dives to go, but the pattern is established. Contrary to conventional wisdom, girls are outperforming boys in math. And also contrary to conventional wisdom, the verbal gap is wide. The verbal results show up on the National Assessment of Educational Progress, but not the math, where boys show an advantage.
Why the difference? I’m not a psychometrician, but my assumption is that when measuring gender gaps, best to go with the assessment that tests every child, versus a test (naep) that’s a sample. When looking at college outcomes (enrollment and graduation), it seems to me that state assessments are proving to be a better predictor than naep. I welcome comment on this from the pros.
Tags: maryland
