Readers will know that I promised a thoughtful follow-up to my previous post in which I criticized DeLisle’s recent Ed Week rant on differentiated instruction.

However, after pondering the subject intensely for a week, I find myself facing a somewhat different and broader question – the baby and bathwater question: why do we constantly fail to distinguish a good idea from confused and ineffective implementations of good ideas, and throw out the idea - instead of refining the policies and practices?

For, surely, this is the issue with DI. The idea could not be more pedagogically and morally correct: design learning to make it most likely that all the varied learners in front of you will learn and be engaged in their learning. Even DeLisle acknowledges the core idea as sound. Indeed, almost every elementary teacher has long differentiated in ELA due to reading level differences, (something apparently unknown to other critics).

But once the going gets rough (e.g. classes are far too diverse; planning becomes more time-consuming), we rise up against “differentiated instruction” instead of tinkering with the way the idea is being implemented.

It’s not just DI. Raise your hand if there has been pushback against UbD, curriculum mapping, block scheduling, authentic assessment, standards-based grading, and problem-based learning in your school. Ok, everyone put their hands down now.

You don’t even have to like these initiatives to see that the implementation problems are rife: failure to think through training and feedback; failure to allocate enough time to experiment with the ideas before full implementation; failure to think through the likely rough spots and misunderstandings of those ideas.

At the implementation level of school reform, it’s one big game of Lucy holding the new-initiative football, and Charlie Brown thinking “THIS time it will work!”

Schools simply do not know how to change themselves. They are status quo machines of the highest order – on par with churches. Worse, administrators – in their naïve enthusiasm and stubbornness to bring change – too often fail to listen to critics or build in self-correcting mechanisms to ensure that implementation can be tweaked all along the way of the reform – as if admitting mistakes in early implementation would discredit the whole idea (and their leadership).

I can speak to this problem with lots of firsthand knowledge related to Understanding by Design over a 15 year period. Let me list a few horrible ways that UbD has been implemented in schools, districts and other countries – without either our blessing or consultation/feedback of any kind:

• In each unit there must be 4 essential questions
• In Year One, every teacher will design and implement all their units in UbD
• Every LESSON will be planned in the UbD Unit Template
• There has to be at least 1 performance task for every lesson
• In Year One, every UbD unit will be placed in the Atlas Rubicon software, with limited PD on either initiative
• Requiring SEPARATE UbD units for each subject at the elementary level in which the one teacher plans and teaches ALL the units.

I could go on, alas, but you get the point.

A process to avoid bad implementation. A solution seems straightforward, based on our sad history in making this mistake:

1. intense study of local needs – identification of gaps between Mission and reality
2. intense study of possible initiatives, given a need/problem statement
3. a process and set of criteria for weighing the pros and cons of possible initiatives/approaches/programs
4. a Purpose statement for the new initiative decided on – the idea and ideal to be safeguarded throughout the work, why the idea suits our needs at this moment in time, and why this is the most promising initiative
5. intense study of past initiatives locally: which succeeded, which failed, and why?
6. policies and criteria for vetting whether a proposed initiative is Mission-appropriate, and whether or not a strategic plan exists that is likely to make the initiative succeed
7. incentives for pioneers to try it out and report back
8. incentives and opportunities for rank-and-file teachers to try out somemanageable aspects of the initiative – thus, some choice for teachers
9. models of exactly what teachers are supposed to do, in their grade/subjects, of the initiative
10. a strategic plan that anticipates and adjusts based on the most likely misunderstandings, concerns, rough spots, and logistical impediments to success. (“This initiative will fail unless we deal with such likely roadblocks as…”
11. a system for ensuring lots of feedback to teachers and sharing by teachers in all scheduled staff meetings, as they try things out
12. a steering committee charged with gathering constant and timely feedback about the initiative and acting on it in a timely way
13. the steering committee recommending key structural changes needed to optimize the long-term success of the initiative, so that all the work is not on the back of individual isolated teachers

The first three points deserve special attention in light of common criticisms of DI – namely, that it is difficult to manage as an individual teacher (True).

DI is one possible solution to excessive i.e. unmanageable heterogeneity in the typical classroom. So, if the problem statement is: too many learners of great difference in ability in certain classes, limiting engagement and achievement for all, then there are additional reforms, beyond DI, that should be considered, too. Maybe we need to reconsider birth-year related grade levels; maybe we need to group more homogeneously (as we happily do in Spanish and upper-level math classes) throughout the period or day. Yes, tracking is bad; that doesn’t mean that intelligent and flexible grouping in classes – especially in a standards-based world where we are accountable for the achievement of all learners – is a bad idea. Then, any proposed solutions might tackle both DI and structural solutions – and be really “owned” by staff since the initiative was a logical response to need rather than a mysterious mandate.

In short, we tend to mandate "solutions" before the problem statement is fully explored, established, and used to consider alternative solutions.

Building an iterative reform system. The title of the post reminds us what tends to happen when leaders fail to do these things. We throw out the baby with the bathwater. i.e. we toss the good idea along with Implementation version 1. (Imagine if software creators gave up after Version 1, and you have some idea of how little we would now value software.)

Resisters/opponents of change get most of their power from the failure of implementation, not sound arguments against the core idea: “See? I told you it wouldn’t work; I told you it was a bad idea.” I think most of the big reform ideas mentioned above are sound, addressing fairly obvious needs for greater personalization, coherence, and accountability. Alas, even “reform” now is a bad word in many quarters (cf. Diane Ravitch) because the implementation of many of good ideas has been so poor.

Nor should we despair over the enormity of the task. We don’t need to be geniuses to change things for the better. We just need to want, solicit, and act on feedback when we initiate any change. That is the key to all modern improvements, from hardware to software to services. Change of any kind, to lead to progress and to llast, involves a robust feedback system. Yet, school-people – be they admins or teachers, be it large-scale school reform or individual experiments in teaching – are prone to charge ahead without an adequate plan, then give up on an idea that doesn’t work out of the box. That’s why it is essential in reform to provide structures and opportunities that send the message: Implementation Version 1.0 is LIKELY to fail. We won’t get this right, most likely until Version 3.5. So, let’s fail early and often (as they say at IDEO) and work to get it right as quickly as possible, based on feedback and advice.

Otherwise, like Charlie Brown, we’ll just be wishin’ and hopin’.