my sister vs. Bill Gates

[quarrel]

My sister is a primary school teacher in rural Pennsylvania, so I asked her her opinion of Bill Gates’s latest experiments at improving education.

Quick summary: over the past dozen years, Gates has given $5 billion — more than half the total charity this cause has received throughout its existence — to scholarships and education grants. He built twenty new schools in big cities with the goal of measuring the effects of smaller classes. He paid for a five-year study to look for common elements among teachers of the most successful classes. He wanted to know if there were teaching practices that consistently work well, and if so, what those are.

Gates discovered a few things. He found that attendance, discipline, and teacher-student interaction all increased noticeably with smaller class sizes, but academic performance and how many students continued on to college rose only slightly. He found no correlation in school performance between states where anyone can be a teacher and states where only union members can. He thinks private schools are excellent and efficient but doesn’t think a voucher program will ever be politically palatable enough to be worth pursuing.

My sister took issue with virtually everything Gates found. First there was his methodology. He tried to test “whether aspects of effective teaching—classroom management, clear objectives, diagnosing and correcting common student errors—can be systematically measured” but he only looked at seven school districts, and all of them urban. That’s not a scientific way to determine what works best in general when most states have a variety of environments and hundreds of districts. (Me? I think it makes perfect sense to isolate variables with a problem this complex.)

Vouchers? Those can only improve American education if private schools are truly better than public ones. They might be, but then again, they might not. The comparisons most often used to support their alleged superiority are confounded by flaws. Public schools must accept all applicants; private schools can and do reject applicants they think will do poorly, even when tuition is not a problem. Private schools also don’t need to give their kids the same standardized proficiency tests as public schools do, so there is no straightforward way to compare the performances of the entire student bodies across the board. You could compare things like SAT scores of those students who try to continue to college, but that’s not a cross-sampling of the whole population. It also relies on the same questionable conception of high school that Gates has: that the primary purpose of high school is preparation for college.

Mainly, though, she is frustrated at the growing trend of holding teachers 100% responsible for children’s performance when a child’s home life accounts for more than 50% of his academic influence. Here my sister brings up the sadly predictable extreme circumstances she’s experienced first-hand: the kids who put half their lunch in their pockets so they’ll have dinner that night; the kindergartener who was locked in his room, beaten, and intentionally underfed the first five years of his life; the seventh grader who shows up in place of her mother at her younger siblings’ teacher conferences because their mom’s in jail and their dad works three jobs. And this is in hicksville, not the inner city where drugs, gangs, and guns make life tough.

Still, I just don’t know. Here’s the thing: I don’t care about the extremes. I’d like to think no sane person would hold a teacher responsible for the poor academic performance of a 15-year-old who falls asleep in class because he works two shifts after school to provide the sole income for a family of three. I’d also like to think no sane person would favorably view a law that forces a school to keep a teacher on the payroll while he is in jail serving out a conviction for molesting his own students. I’m talking about the everyday ordinary cases, not the outliers. I simply find it implausible that teacher proficiency cannot — just cannot, at all — be evaluated in a sufficiently accurate, uniform, and objective manner that we know who deserves a pay raise and who needs more training or a career change. I know teacher skill is far from the only factor in determining a kid’s performance, and could easily be a minority influence, but teachers do vary in skill, right? That’s true in all other occupations. And why can’t we adjust for things that aren’t their fault? Firefighters in drought conditions in the rural southwest aren’t expected to keep fires as contained as those in Baltimore in the spring.

Comments

[shaterri]

I simply find it implausible that teacher proficiency cannot — just cannot, at all — be evaluated in a sufficiently accurate, uniform, and objective manner that we know who deserves a pay raise and who needs more training or a career change.

I'm honestly not sure. As you say, the outliers are easy, but when it comes to folks in the middle, and especially relative comparisons between them, things get a lot fuzzier, in ways that have nothing to do with teaching specifically but just with the multidimensionality of skills in general. Baseball evaluation is sort of the classic example of this; if player A's slash line (Batting Average/On-base percentage/Slugging) is .268/.374/.502 and player B's is .326/.376/.493, which one is better? Does it matter if player A strikes out twice as often as player B? What about the difference in position, if B plays a position where offense is at a premium (say, shortstop) and A plays a position where offense is easily found (e.g., first base), should that matter? What if player B's 'batting average on balls in play' is 50 points higher than median (and we think that players have no or limited control over BABIP), should we expect that to regress and should we hold it against him? I'm a firm believer in objective measurements and think there's a lot more that we could be doing to measure both teachers and students, but I think we'll need a lot more understanding of those measures and what they really mean before we can 'safely' make critical policy or hiring decisions using them.