Claus von Zastrow, Executive Director of Learning First Alliance, rightly points out in a blog post that the failure of class size research to show enough test score gains to be worth the cost past 3rd grade really points to the failure of test scores themselves to tell us much.

Interestingly, the author of the first comment speaks to precisely my train of thought when I looked into the class size research and its limitations. I ended up looking beyond just class size to the other–intertwined–relational factors. Tom came across my book in that process.  His comments are:

California Watch, an investigative journalism group, recently studied class size in the 30 largest districts in California to see what was happening with the voluntary 20-student cap in K-3. This map lets you see exactly who’s opting in and who’s opting out.

Here’s a good recent piece on class size as it impacts African-American students. What’s brought out in the article is also that Californians pay for the (now crumbling) K-3 reductions to 20 with the largest middle and high school classes in the country.

John Rogers, director of UCLA’s Institute for Democracy, Education, and Access (IDEA), said … it’s important to note that California already has the largest class sizes in the country in middle and high schools, particularly in the subject areas of English, social studies and science.

Besides cost, the most common argument against smaller classes is that it would lower teacher quality by bringing in new teachers. In my book I offer several refutations of this. My main response is that new teachers brought on through class size reduction will be instantly better because the job will be easier and the teaching force will stay better because frustration attrition will go down.

Another argument I brought up is that the percentage of K-12 teachers per capita approximately doubled in response to the baby boom and then stayed steady since its end. And yet I’ve never seen anyone try to argue that teaching got worse between 1950 and 1970. Here’s my source: http://www.bls.gov/opub/mlr/2006/03/art3full.pdf

Indeed I point out in the book that the teaching force simultaneously became more likely to hold graduate degrees during that same time span.

This brings us to the issue of higher ed, which is my main point here:

It struck me recently that another educational trend could be scapegoated by this same logic that assumes there are a finite number of smart-enough/educated-enough people available to fill teaching posts. What of the trend to send more people to college? Couldn’t that be a significant brain-drain on K-12 teaching if more teachers are needed in higher ed? I went back to the same Bureau of Labor Statistics report and found that college and university teachers grew constantly as a percentage of the population over the last century. Since 1960 it appears to have tripled, and the trend has shown an acceleration since 1990.

Couldn’t this be one place where the coveted “best and the brightest” potential K-12 teachers are going? The ones who are so good that class size doesn’t matter? Yet often the same people who argue that we will dilute the quality of the K-12 teaching force if we try class size reduction are also those arguing that we must send still more of our workforce to college to be ready for “the information jobs of the future.” Those aims might be in competition.

Add to that this question: Couldn’t teaching be one of those information jobs? If we are sending an increasing percentage of our students to college, why is it supposed that there is some finite and ungrowable number of people smart-enough/educated-enough to be teachers? Either the initial assumption of a finite supply is wrong, or higher education actually does nothing to make people smarter or more educated, which would leave the goals of intensifying K-12 education and universalizing K-16 both pointless endeavors.

The definitely non-pointless endeavor would be not to neglect our kids with large classes, to do education more ethically—rather than race off into fantasies of the infinite perfectibility of the human being, which at bottom reveal their belief that people are currently stupid and will stay that way until “saved.”

All Things Considered’s recent segment on class size perhaps indicates that the issue has leaped from the local interest pages to the national media stage. However, the bigger the stage, it may be that the more depersonalized, the more test-scorey the analysis tends to be. The local pages often frame the issue as local parents being concerned their kids will get ignored or local teachers will be overwhelmed. As described in my last post, K-12 education remains an immanent experience not necessarily reducible to its results on later life.

Jeremy Finn could have done a better job in the interview of not sticking merely to K-3 evidence of class size reduction’s value, avoiding the “youngest kids only” frame. After all, the story was introduced with the anecdote of a high school class with 43 students, which he called “Too big, period.” But Finn did do a great job avoiding the “just the test scores ma’am” frame.

There’s just a point at which teachers can’t really get to know the individual students, can’t teach the materials. Students split off into splinter groups and withdraw from the classroom interactions, and 43 is well into that range.

As a counterargument, Madeleine Brand references a New Yorker article by Malcolm Gladwell:

He writes: “You have to cut the average class almost in half to get the same boost that you’d get if you switched from an average teacher to a teacher in the 85th percentile.” So in essence, he’s saying that teacher competency is better and more important to student achievement than class size.

I looked up the reference and found that it’s an echo of the usual suspect Eric Hanushek of Stanford’s Hoover Institution. Same-old-same-old teacher quality vs. class size. In my book I argue that class size reduction would help attract and retain better teachers while at the same time making the not-as-good teachers more effective.

After years of worrying about issues like school funding levels, class size, and curriculum design, many reformers have come to the conclusion that nothing matters more than finding people with the potential to be great teachers.

Matters for what? For raising the average in the abstract for the purposes of symbolic progress? For American power? Or for for equity and child rearing? (Again, see my last post.) In general, teacher quality folks lie when they claim there isn’t a focus on teacher quality in this country. Teacher ed programs focus on it, administrative interviews and evaluations focus on it, ongoing professional development focuses on it. What they are fabricating is a Lake Wobegon fantasy of above averageness in teachers and innocuously calling it a focus on teacher quality. It’s pure and simple teacher bashing because it implies they–we–are hopeless raw material. You know you’re doomed when you start here: “If we simply found an elite group of saviors…”

But when I read the rest of the article, a more complex picture emerged that isn’t at all focused on test scores like Hanushek: It’s focused on interpersonal interaction, making students feel listened to and feel that adult attention is available, i.e., what my book is about. Examples:

Among [key teacher skills] is “regard for student perspective”; that is, a teacher’s knack for allowing students some flexibility in how they become engaged in the classroom.

“Almost every time a child says something, she responds to it, which is what we describe as teacher sensitivity,” Hamre said.

It was a key moment. Of all the teacher elements analyzed by the Virginia group, feedback—a direct, personal response by a teacher to a specific statement by a student—seems to be most closely linked to academic success. Not only did the teacher catch the “Me!” amid the wiggling and tumult; she addressed it directly. [puctuation fixed]

“Mind you, that’s not great feedback,” Hamre said. “High-quality feedback is where there is a back-and-forth exchange to get a deeper understanding.”

The preschool teacher with the alphabet book was sensitive to her students’ needs and knew how to let the two girls on the right wiggle and squirm without disrupting the rest of the students; the trigonometry teacher knew how to complete a circuit of his classroom in two and a half minutes and make everyone feel as if he or she were getting his personal attention. But these aren’t cognitive skills.

But Gladwell discusses these as skills in a way that assumes some got ‘em and some don’t and probably won’t ever or won’t get ‘em fast enough. Aren’t these also functions of class size? Doesn’t this analysis amount to stating that the key teacher skill is the ability to make a large class feel like a small class? To make teaching feel more like tutoring?

And isn’t Gladwell’s logic analogous to Hanushek’s but with more ethical criteria? Both basically argue, “Let’s assume we won’t get smaller classes. That means we need teachers (seen as a static set of skills) with more inborn skills to make education improve.” Neither acknowledges that smaller classes would logically make it more likely that those who lack key skills initially would acquire them in the first couple years. Made not born. Not only because the classroom disorder stakes are lower and teachers are less likely to resort to counterproductive control strategies, but also that they’re simply less likely to leave the profession before they get good.

This poor logic suffers from the problem of ignoring the passage of time, ignoring the fact that teachers usually start weak and end strong. It sees things synchronically rather than diachronically. Those getting the worst test scores are often taught by those with the least experience. So we’re not in the easier-to-solve situation where some students are simply stuck with a teacher who will never be good, but that teachers who are (or have become) good have the bargaining power to switch to schools (or assignments within the same school) where the students are easier to teach anyway. Plus those students are generally whiter and richer, more like the average teacher culturally. Good teachers do not magically escape the classism and White Supremacy that permeate our society.

Suppose that Mrs. Brown and Mr. Smith both teach a classroom of third graders who score at the fiftieth percentile on math and reading tests on the first day of school, in September. When the students are retested, in June, Mrs. Brown’s class scores at the seventieth percentile, while Mr. Smith’s students have fallen to the fortieth percentile.

This artificial setup suggests a false problem in education: that poor teaching is doing a disservice to otherwise similar, interchangeable kids, doing the average kid wrong. This is a reiteration of the assertion that the average is too low, that raising the average will get us something that isn’t merely symbolic. It is not focused on equity or child rearing (the goals I argue for in my previous post).

This is the tyranny of the average, the fantasy encapsulated by Garrison Keilor’s refrain about his fictitious hometown of Lake Wobegon, “where all the children are above average.” Hanushek is at least honest about thinking in terms of percentiles and averages. He wants to raise the average quality of teachers by measuring them against a standard and kicking out those below the average.

It’s criterion-referenced testing (versus norm-referenced) that is dishonest about subscribing to the Lake Wobegon fantasy. How are we to imagine the criterion of a high-stakes test or “grade-level proficiency” are arrived at? A criterion is clearly constructed rather than given by divine revelation. And who will disagree that it isn’t ultimately based on the average of what current students can do that’s easily and cheaply measured en masse? A criterion is the average we wish everyone could be above. It is usually set a bit below the average, of course, so that the fantasy of having everyone meet it is a tad less farcical.

The desire to raise the average by setting a magical criterion might be an acceptable and useful fantasy if it didn’t usually include the desire for infinite progress. Behind every argument that our current standards are not good enough, or are falling, is always the belief that stasis is retrogression, that only progress is non-retrogression. This thinking clearly underlies NCLB. It is the logic of scarcity I referred to in the last post.

Gladwell’s solution is the same as Hanushek’s despite replacing test scores with attentive, compassionate teaching: weed.

we shouldn’t be raising standards. We should be lowering them, because there is no point in raising standards if standards don’t track with what we care about.

This is the good point he quickly betrays by simply creating a fierce division between what students need and what teachers need–students need attentive help and teachers need surveillance. (I worked in a school with this same fierce division. Those who left before me were characterized as having subordinating student needs to their own, of not caring enough about the students, of preferring richer and whiter students. I assume that’s how my departure was also understood.)

Teaching should be open to anyone with a pulse and a college degree—and teachers should be judged after they have started their jobs, not before.

Hanushek’s means at least resembles his ends: Standardize the teachers by picking the ones best at standardizing children.  With Gladwell, the teaching of teachers would in no way resemble the unstandardized teaching we’re looking for in those teachers.

Gladwell wants tenure to be harder to get and he wants merit pay, because “the only way to get people to try out for what will suddenly be a high-risk profession is to offer those who survive the winnowing a healthy reward.” Thus he wants cutthroat teachers focused on external rewards for themselves but who are also somehow simultaneously and extra-ordinarily focused on (appearing to be?) sincerely giving interpersonal feedback to unique kids in the here-and-now of the classroom, to instilling in kids the intrinsic joys of learning and that school is not high-risk and cutthroat in the way that very teacher’s job is.

Gladwell thinks we need “an apprenticeship system that allows candidates to be rigorously evaluated.” In other words, we don’t need an apprenticeship system because it will better impart skills like giving rich feedback–by giving those teachers-to-be richer feedback, incidentally–but because it will submit them to better surveillance to determine more definitively whether they were born with the right skills–or acquire those skills quickly enough not to require good teaching themselves.

In yet other words, you as a potential teacher have to be born with the ability to show students that they are teachable no matter what they’re born with, but these skill can’t be taught to you–because you are a teacher now, not a student, something completely different. That is the fierce division. Gladwell would clearly fire a teacher of children if they operated on the philosophy of teaching he is asking teachers/bosses of teachers to operate on.

But there is a positive flipside to what Gladwell is arguing here: If you slice off the surveillance and rejection tenor of what Gladwell wants in teacher training, he’s suggesting that we move teacher education in the direction of practice rather than preparation, as do the many who argue that we move it toward an apprenticeship model. He is arguing that there is a certain folly to the philosophy that education as preparation offers any real guarantees for life after education. He’s arguing that teachers learn how to teach (and that they’d better learn quick!) in the practice of teaching, not in the teacher-ed classroom.

This also necessariliy suggests something about education by those teachers: that we move it from preparation to practice, to making schooling a cultural activity worth doing in its own right. Gladwell unintentionally suggests that we consider K-12 education as a form a childrearing, not a preparation for the supposedly real form of living, i.e., adulthood.

This is what I was getting at in my last post with the idea of the 13 years of schooling as an experience that has inherent value, not only value via its results. Dewey said as much 100 years ago: Education as preparation devalues itself into a 13-year prison.

Ted Sizer and the Coalition for Essential Schools often take the focus off of school size per se and put it on pupil load, or what a new book by William Ouchi calls “total student load”–the total number of students teachers have in their current classes.  CES says 80 should be the max. They summarize their goals as:

• Personalized instruction to address individual needs and interests;
• Small schools and classrooms, where teachers and student know each other well and work in an atmosphere of trust and high expectations;
• Multiple assessments based on performance of authentic tasks;
• Democratic and equitable school policies and practice;
• Close community partnerships. (source)

In The Secret of TSL [total student load]: The Revolutionary Discovery that Raises School Performance, fresh off the presses September 1st,  Ouchi acknowledges Sizer and CES’s focus on pupil load and incorporates it into a a theory of district management: Empower principals with the freedom to make more budgetary decisions and they are likely to choose to reduce total pupil load on their teachers (partly by hiring more), which is likely to improve outcomes.

When I sought in writing my book to link class size, school size and the length of time students and teachers stay together, I broadened this idea of pupil or total student load into relationship load. This offered a concept that could

a) clearly communicate the import of total student load

b) act as an umbrella to include class size, school size and length of relationship

c) include the perspective of the student and the parent or guardian as also bearing a load of school relationships that affects their “performance”

d) refocus on the broader concern of schooling as a part of rather than apart from child rearing

So though I applaud Ouchi’s analysis, I think it’s still framed within a conception of schools as “apart from.” His other educational management book is Making Schools Work, whose title echoes an entire discourse around the mantra of using research to find out “what works” in education. “Works what?” is the question we have to remember to ask so as to keep an eye on the assumptions behind all this work.

“What works” is code for what can be shown in random sampling to produce quantified snippets that symbolize educatedness, which in turn loosely prophesies employability or usefulness to economic growth, which in turn signifies American superpowerhood.

What can’t be shown to be true in general (because it isn’t general and never will be) is what works with unique people but not with other unique people, so this educational knowledge that only materializes in specific classrooms is not classified as “what works.” And what works to, say, not neglect students while educating them, doesn’t necessarily lend itself to quantification, thus doesn’t get labeled “what works.” (See my post comparing small classes to crosswalks. They aren’t meant to make traffic more efficient–they’re meant to make traffic more ethical, to correct for the power imbalance between cars and pedestrians.)

But it is possible to use quantitative research to investigate ethical improvement. You can measure, for example, whether crosswalks efficiently improve the ethics of car power. Analogous work gets done in education. It’s not the method per se that’s the problem–it’s the culture of assumptions that surrounds the method and those initiated into it. It’s a means-end question. “Raising school performance” is a questionable end, as is its cloaked sister goal of American economic hegemony. Money is an instrument for more important ends.

If relationship load is an instrument, it’s an instrument only for relationship. If total student load is an instrument, then, it isn’t for “raising school performance” but literally for the total student, the whole child. The Secret of TSL: The Revolutionary Discovery that Raises School Performance. The real secret is not its transcendent instrumentality but its inherent, immanent value.

“Raising school performance” in the abstract or across the board is valueless. It only makes sense within a frame in which there will never be enough–a deep belief in scarcity, a religion of “we are never good enough.” It only makes sense within a competitive frame where raising school the performance of my kid or my nation-economy  has value because I’m assuming others will not get access to this “raise.” More ethical values are (a)  raising school performance where it results in equity across lines of social stigma and (b) raising children.

The complication to how I value equity in and by education is that I don’t believe schooling is necessarily the alchemist’s stone of equity it’s trumped as. Oftentimes it merely serves as a rationalization for inequity because it offered the have-nots their chance, so they must’ve blown it–they more clearly deserve it than the serfs of old.

Don’t get me wrong, there’s no harm in trying for equity in and with schools, but there is harm in doing this in lieu of fixing what’s more endemic to education itself. This would be trying for what’s farther off without getting what’s close and easy: more ethical raising of children. In other words, let’s not sacrifice the definite gains to all children/future adults of a less mass-produced upbringing on the altar of better test scores for kids of color and poor kids. Because there’s no guarantee–much as I’d like there to be–that better test scores will make much of a dent in what holds the marginalized back after graduation.

The goals outlined in my book are thus to help fix primarily (but not exclusively) the problems specifically of schools (unethical relationship loads), and secondarily the problems also manifested in schools (other social justice concerns). Luckily it’s not really an either-or choice because relationship load reduction has been shown to equalize outcomes across social justice fault lines.

Ultimately, though, I could take or leave schools as an instrument as well–keep them or scrap them entirely–so long as I got to my ends-in-themselves of equity and child rearing.

And that’s the best I’ve come up with.

Mike Klonsky of the Small Schools Workshop runs a good education blog, Small Talk, whose “small” refers to school size, but he commented recently on small classes.

For more than a decade now, I’ve been hearing from right-wing think-tankers, like Fordham’s Checker Finn, Mike Petrilli, and the Hoover Institute’s Eric Hanushek, that class size doesn’t matter. The research, they claim, shows little benefit to student achievement scores by reducing the number of students in a class. With class sizes in L.A. high schools soaring to 50 per class and Brooklyn kids being squeezed into classrooms with a shoehorn, Finn and his crew have suddenly grown quiet on the topic.

Two comments:

a) It’s true that the media are actually noticing class size for once. NPR’s All Things Considered ran a piece on Thursday.

b) If we didn’t keep letting the likes of Finn and Hanushek set the frame for what research ought to be about (test scores, achievement narrowly defined) fewer might have listened to them thus far.

This is a similar problem to what led the waning of interest in the small schools movement. Klonsky rightly criticizes how “ownership society” discourses corrupted the small schools movement (example). But so did test score pipe dreams it couldn’t fulfill.

If this is a nascent small classes movement we’re seeing, let’s not poison it with the same faulty goals. Read my book for what I mean by better goals.

The photo Mike borrows from the LA Times shows two kids sitting on a filing cabinet at the back of the room. There’s an irony to the photo. One of the things I noticed in my research for the book was a pattern in the cover photos on education books. When they picture students it is never anywhere close to an average class-size-worth. It is always an ethically sized group of kids. If a whole class were depicted on the cover it would depersonalize the kids, make them look neglected, unhappy; make the reader feel uncomfortable or perhaps disinterested. The gist of these observations is that what is true with cover photos seems also to be true with the daily lives of the students and teachers they depict.

The irony is that even this photo pictures only eight visible kids and only four faces. It’s the filing cabinet that does all the work. There’s still something taboo or ineffectual about actually taking a wide angle shot of the 50+ kids. Somehow you just don’t care about a group of that many people they way you care about someone close up. And that’s the whole point, ain’t it?

In my book I reference the class size research of Peter Blatchford as an example of research too focused on test scores:

Far too many studies seem to assume (perhaps rightly) that policymakers will snicker at discussions of attitude, morale, or emotional well-being. Hence, most research on class size reduction centers unflinchingly on test scores. Nonetheless, Blatchford himself noted, “It may be that a smaller class size allows a teacher to approach children in a more personalized and more humanistic way, but this would be hard to measure and then enter into statistical analysis!” (2003, p. 160). The exclamation point is his. Other researchers have made similar comments as postscripts to their studies, for example:

It should be noted that these enrichments to the curriculum may not be reflected in evaluations of class-size effects. Typically, achievement tests in reading and mathematics are used to evaluate outcomes. This narrow definition of achievement overlooks learning in other academic areas, areas which may be valued by consumers of education. Also, many of the enrichment areas are intended to promote positive attitudes, enthusiasm, and overall learning skills. These factors may have long-term effects not in evidence on short-term achievement tests. Research in education may be misled by its focus on short-term achievement outcomes. (Cahen et al., 1983, p. 206)

Nevertheless, most research continues to be misled in this way and to make only minor acknowledgments of the possibility. (pages 123-124)

Kudos to Peter Blatchford! Turns out he has followed up on this acknowledgment (since my book researched had stopped) and began to do class size research that looks inside the black box rather than simply looking at the easily measurable outputs from that box.

There has been a tendency to assume that links between class size and academic attainment are primary, and that links with classroom interactions are secondary. But it is important that pupils engage in work in classrooms and important that interactions with teachers are work focussed and where possible individualised, irrespective of whether we can show that these obviously cause changes in measured test scores. The quantity and type of interactions with teachers, and the degree of attention in class, are therefore important ‘outcomes’ in their own right.

(Do low attaining and younger students benefit most from small classes? Results from a systematic observation study of class size effects on pupil classroom engagement and teacher pupil interaction Peter Blatchford, Paul Bassett and Penelope Brown, 2008,  p. 24)

Let me note from the outset that the subjects were British and that there is no disaggregation by race or other possibly significant cultural factors. The paper finds some common sense results like the fact that individual interactions with the teach skyrocket (exponentially):

Similar results are found for pupils being the focus of the teacher’s attention. (There’s a book about that!) Meanwhile, time on task rises more linearly:

When the data is disaggregated by student achievement level, there are some stunning results suggesting that it is the low and mid achieving students who need the teacher’s attention (as indicated by off task behavior) and the highest achieving that seem to just need the teacher to get out of their way. The irony is, of course, that the highest achieving students are often tracked into (or their parents pay their way into) exclusive programs and schools with smaller class sizes!

elementary:

secondary:

Similar graphs depict how likely it is that teachers spend time on negative behavior with students from each achievement level. (Curiously, the researchers do not report the aggregate correlation between class size and their “teacher deal with negative behaviour” coding.) That high and medium actually go down with larger class size might reflect that teachers simply opt to let higher achieving kids get away with more and focus on reigning in the lower achieving kids. I describe this in the book as a process where in larger classes teachers tend to divide the class into allies and adversaries. The researchers write,

We feel that these results on on and off task behaviour are significant because they show that the problem of large classes, especially in older pupils, is particularly marked for the pupils who are already attaining at lower levels, and that it, in turn, is associated with teachers seeking to control this kind of behaviour. It is easy to see how these two kinds of behaviours can become self reinforcing, exacerbating each other and making the situation worse. In contrast, smaller classes seem to allow an environment in which low attainers are not only less off task but are less likely to receive corrective talk from their teachers. This appears to be a more productive educational environment. (23)

And then there are results that are ambiguous in import. First, there is less student-to-student interaction as classes get smaller, whether its on or off task interaction. “It seems likely that with more pupils in the classroom there is less of the teacher’s individual attention available and this presumably means that pupils come to interact more with each other instead, in both acceptable and unacceptable ways” (23). My response to this would be that (a) we probably need to remind and train teachers to keep doing group work in small classes, and (b) we need to make sure not to glamorize peer cooperation that is out of balance with adult attention. (See chapter 4 of my book.)

Second, teachers appear to spend less time actually “teaching” the class in the traditional sense. According to the researchers, “in smaller classes pupils get more individual attention, while in larger classes they spend more time listening to the teacher talk to the whole class. They are perhaps getting more educational input, but this is at the expense of it being largely passive and as part of a large group” (22).

elementary:

secondary:

This could perhaps explain why the test scores don’t skyrocket in the same way other indicators do. Perhaps teachers can accomplish the same amount more efficiently and don’t necessarily feel the need to make small classes learn more just because they could. The “more! better! faster!” discourses we have been bombarded with would say these teachers are being lazy, are squandering the opportunity to beat the [fill in competitors here]. I plan to focus my doctoral research around questioning this response. Why more? Why better? Why faster? Instead, I want to propose that more equal is the only real urgency. And Blatchford, Bassett and Brown’s research suggests that more equal (as opposed to better than)  is precisely the tendency inherent in class size reduction.