Beware of Dubious Teaching Secrets

  • Share
  • Read Later
One of the most emailed stories on the Web last week reported that junior-high boy students do better when taught by a man, and girl students do better when taught by a woman. Here was a magic-bullet explanation for The Boy Crisis from Stanford's Hoover Institute: the boys are lacking because 4 out of 5 teachers are women.

We weren't surprised this story ran absolutely everywhere, from USA Today to the Hindustan Times — who doesn't enjoy debating whether boys or girls are weaker? It seemed fitting that the story landed the same day that Billie Jean King, victor over Bobby Riggs in the 1973 "Battle of the Sexes," was crowned with her name on the National Tennis Center at Flushing Meadows.

However, the media overlooked that the study's underlying test data was from 1988, meaning this tiny skew toward a teacher's sex was evident among 8th graders almost 20 years ago. We have no idea if it's still true today. Nor could the study's author, Thomas Dee, convincingly explain why this "teacher gender effect" appeared in science class but disappeared in math class.

But the biggest failure of reporting was that no one insisted on an answer to a simple question: "How much better do the students actually do?"

Not much better at all, according to fine print of Dee's study. On a simple 100-point reading test or math test, this "teacher gender effect" might alter a boy's score by one or two percentage points. Hardly the difference between a C and a B.

We showed the study to Leslie Scott, the Principal Research Analyst at the American Institutes for Research. Unimpressed, she asked, "This may be statistically significant, but is it substantively significant?"

You would be amazed at the weird variables that education researchers have found can improve student performance. They might all be factually proven, but that doesn't make them meaningful policy fodder. For instance, classroom windows. Studies have found that math scores improve when classrooms have windows — especially if the window looks out on verdant lawns. This is no meaningless improvement; it's as powerful a factor as whether computers are in the classroom. In one Fresno, Calif., study it even mattered which direction the windows were facing! (Facing east was best). But we'll never see two politicians on Meet the Press arguing over window-installation appropriations.

What else boosts math scores? Well, don't laugh, but teaching on a whiteboard with dry-erase markers is better than blackboards with chalk. Students can simply see better. Another study showed that humming fluorescent lights are particularly hard on students — they make gifted students test "ungifted." And light blue paint on the walls of a kindergarten lowered misbehavior by 22%.

Our point isn't that wall color deserves to be given equal due in the headlines. Rather, we just want the media to recognize when it's being fed a carefully timed and calculated diversion. We are as easily distracted by gender politics as students are by those fluorescent lights. Schools are opening around the country this week, and people are talking about teacher gender rather than teacher quality.

Even what your child eats for lunch matters more than the sex of his teacher. In fact, getting good calories into students has so consistently been shown to improve test scores that some schools in Virginia have been accused of cheating a la Barry Bonds — by juicing their students on test days. On those days, the free lunches that poor students received had unusually higher calorie counts. Among those students, English test scores rose by 14%, and 18% in History and Social Studies. Maybe they scored higher on the next round of child obesity tests as well.