Today’s Puzzle: What Do We Know?
Posted: December 13th, 2012 | Author: Michael Goldstein | | 5 Comments »
Do you think we mostly know “what works” in education?
Or
do you think we mostly do NOT know what works?
I.e., do we know what to do, but lack the willpower, or resources, or something? If you, Dear Reader, were “czar of everything” in American K-12, do you think you could get a much better result than what we have today? Then you’re in the first camp.
If you think something like this — “Gee, we now do a lot of things that defy common sense. I know those are dumb. However, even if you cleared out that underbrush, the low-hanging fruit of stupid things, I still wouldn’t be clear on how we’d get to a place where kids are doing much, much better than they are today” — well then you’re in the second camp.
Many of my friends and colleagues who work in K-12 believe the first. I’m not sure if it’s the majority. But I’d guess it is.
I’m in the second camp.
A couple weeks ago, Match hosted Roland Fryer to lead a Friday night conversation. He’s pictured above with my colleague Stig. Here are some snacks we served. The fancy kind. No Doritos or little hot dogs. So I abstained.
Roland discussed his work at EdLabs. He leans towards the “mostly we do NOT know what works” view of K-12. So his team is tackling a whole bunch of edu-puzzles. A lot of his ideas try to get “outside the normal box.” That is, to achieve breakthroughs, you need to try out unusual ideas.
He made an observation that resonated. In medicine, if you try an unusual idea, and it doesn’t work, you kind of shrug, you know it was a longshot, you move on. In venture capital, if you try 10 companies, you’re assuming most will fail, you’re hoping one hits big. In K-12, if you try an unusual idea, the professional culture is often that people “hang their heads” when it doesn’t work. The idea of failure as a necessary way to achieve breakthroughs is not part of the typical K-12 culture. Roland: “Whereas my view is, hey we just learned something.”
One of the attendees was former Tufts president Larry Bacow. He’s now “president-in-residence” over at Harvard Graduate School of Education, and he now lives a few blocks away from the Match high school. Larry is pictured here, over what I believe is the shoulder of Antonio Gutierrez, a Match grad and a Union College grad who now works with Alan on our work in Lawrence.
He said something along these lines, and I’m paraphrasing:
It strikes me that higher ed is in some ways the opposite of K-12 in this country. Our higher ed is often described as the envy of the world; that is not true of K-12. Also, our higher ed has enormous variation, customer choice, and money following the customer. Our K-12 system is the opposite. I wonder if there’s a link between the success of higher ed and the structure of higher ed….
I followed up with him to ask if it were okay to post his remarks. He agreed, and added:
You understood precisely what I said. But there was more. Higher ed succeeds because there is competition for students, faculty, staff, resources and mind share. This competition breeds innovation across the sector. Students are mobile (as are all the other resources) so both innovation and success are rewarded. By contrast, in much of the rest of the world higher ed is centralized with little competition as in the US. As a result, you see little curricular innovation and excellence is not rewarded.
Now let’s look at K-12 in the US and abroad. In the US, we decentralize funding of K-12 education largely to the municipality given our dependence on the property tax to fund K-12 education. Students are not mobile (unless their families have the resources to move to a better school district. The fact that many do for precisely this reason underscores the value that people place on good schools.) We wind up with huge disparities in the quality of education based strictly on the wealth of school districts.
By contrast, in the rest of the world, funding of K-12 education is centralized and curriculum is also standardized. You have higher levels of achievement in part because resources are more flatly distributed across students geographically. Moreover, because of the concentration of funding, you can actually support curricular innovation and experimentation if you choose to do so which is tough to do at the district level in a disaggregated system with large local funding disparities.
This is the first time I have written any of this down. I don’t profess to be an expert on k-12 education. These are musings and only that. Please feel free to correct any misperceptions that I have.
I thought it was provocative. If I understand, Bacow is saying our K-12 is the worst of both worlds: decentralization without the bottom-up innovation benefits that come from true competition, and without the top-down R&D benefits that can come from centralization.
EdLabs has a bunch of interesting stuff going on. They’re trying out some Lemov train-the-trainer in Texas. They’re working with teachers and the principal at a Denver school on an unusual high-dosage tutoring model. Roland’s got a fascinating paper coming in a few weeks from Harlem Childrens Zone; it will get some attention.
I think one important framing question here is “Does a truly successful educational system, or even small group of schools exist? Or has one ever existed?” There are arguments for “no,” but it is important to understand if we’re taking a Utopian frame here.
I think, for example, that Joel Klein thinks he is on the way to creating an educational system greater than has ever been seen on Earth. I also think he’s completely full of it.
I tend to think what you and Fryer are actually saying is “We don’t know anything about creating a successful educational system for all students in a profoundly unequal and segregated society.” And I think that is because it is an impossible task, and moderately immoral in its enabling of the larger status quo.
I don’t agree with Bacow about the relative success of higher ed, at least in terms of teaching and learning. There’s not much evidence for that. On the other hand I do agree that our particular mix of centralization and de-centralization is particularly unhelpful.
Hi Tom,
Sports analogy. And you know how it went last time (Papi!) I went down that road.
But bear with me….
Gronk broke his arm. Is the best guess that it will take 2 weeks, 6 weeks, or 15 weeks to heal?
If I know the answer, I can take action based off that knowledge, in terms of game planning or acquiring other receivers.
Seems like a fairly straightforward thing to “know.”
And indeed, we know. The doctor can say “Typically that type of break will take about 6 to 8 weeks to heal.”
* * *
I’ve got a 3rd grader who can’t yet add, subtract, divide, multiply 2-digit numbers. Assume no special needs, no cognitive issues.
Roughly how much 1:1 tutoring is likely to solve that, get her to grade level?
If the typical answer is 10 hours, 40 hours, or 200 hours — any of those would be valuable information, even as I understand the limits of “typical.”
Geez, if it’s 10, let’s get that solved over Xmas break. We can bang it out right now.
If it’s 200, somehow I probably need to deliver an hour a day for a whole school year….
But I don’t even have a PARAMETER for what it will take to do the job. Do you have a good estimate? Does our field know?
Thanks for sharing Larry’s thoughts. My wife’s a professor and Tufts and loves Larry – he did a great job there.
Like Tom, I agree with him on the unfortunate mix of what’s centralized and what’s decentralized. Funding and accountability would ideally be centralized, with school control more decentralized.
But I’ve been spending a bunch of time looking into higher ed in more depth recently, and this ‘envy of the world’ trope about our higher ed system is starting to bug me. Do we know what works in higher ed — as regards student learning — any more than we do at the K12 level? Like you, I’m in the no camp.
There’s no doubt our top hundred colleges & universities are the envy of the world. That’s mostly based on the quality of the research at those places and the value of attending school with a lot of other smart students who are on a path to success. I think we know what it takes to produce a great research university or selective liberal arts college ($ is vaguely involved).
But do we know what it takes to produce great student learning — and even to ensure that kids actually finish college? Looking at our full set of 4500 colleges rather than the top 100, we have essentially NO idea whether and what students are learning. Graduation rates are middling at best and horrible for lower-income students. What little research has been done on learning outcomes is depressing. And yet price inflation is healthcare-like and students come out saddled with debt.
It’s really interesting when you look at it structurally. I totally share Larry’s bias toward designing competition, choice and diversity (of models/offerings) into a school system. So why isn’t that competition creating better outcomes?
Like any other market, competition on its own doesn’t get you there. In the absence of meaningful accountability, transparency or regulation, colleges aren’t competing on the right set of things. Right now, competition among 4-year colleges revolves around US News & WR rankings and ‘Carnegie climbing.’ It makes college more expensive without improving learning or graduation outcomes, which barely figure in the criteria. Disruptive tech-enabled innovation will address some of this by creating lower-cost options, but will it necessarily produce better learning outcomes or graduation rates if market structure stays the same? I’m a long-term technology optimist so I think the answer it that it may help, but new models won’t solve our problems on their own.
So I’m not sure that we’ve got the particular mix of centralization and decentralization right even for higher ed. It’s at least one big step better in that it’s premised on choice and competition. But funding equity is worse than K12 (even with Pell etc.); there’s minimal centrally-imposed transparency or accountability (worse than K12, where there’s cleaner graduation/persistence data and at least some minimal definition of learning outcomes headed in a better direction w/Common Core); and the state of subsidized Roland-like research to learn what works is similarly pitiful.
Your summary of Larry’s take on K12 seems to apply to higher ed with some slight tweaks: decentralization without the benefits of bottoms-up innovation that come from true competition around learning outcomes & graduation rates; and without the top-down R&D and equity benefits that can come from centralization. Curious whether others would agree?
[...] shared a comment on Mike Goldstein’s interesting blog post here talking about whether we know what works in K12 ed and comparing K12 to higher ed in the U.S. Mike [...]
[...] http://www.startinganedschool.org/2012/12/13/todays-puzzle-what-do-we-know/ [...]