This is a final thought triggered by the Dell Foundation’s Blended Learning Report, although it applies to many reports, presentations, and discussions about blended learning.

The word “model” is commonly used when the idea being presented is more appropriate to apply to a school, not to a model. An example is this sentence from the Dell report:

“Do students in blended learning models show changes in academic achievement that differ significantly from their peers’?”

A “model” may be a representation of something in the real world, or it may be an exemplary system. In the blended learning field the common models are those described by the Christensen Institute, such as the rotational model. But as the researchers at Christensen have stated, their models don’t represent exactly what’s happening in most schools. There are many versions of each model, and the extent to which the blended learning is well implemented in terms of providing the necessary professional development, devices, bandwidth, etc. impacts student achievement much more than the model.

At times researchers and policymakers refer to the “fidelity” of the blended learning implementation, with fidelity perhaps meaning “the degree to which something matches or copies something else.” But that’s not quite right either, because educators who are implementing blended learning know that they have to tweak the general models to make them work in their existing schools. This tweaking to make the instructional approaches work is a feature of good blended learning, not a bug. A startup school may be able to look at the blended learning guides that are available, choose one, and implement something that looks very much like the guide. Existing schools are never going to be able to do that. They are working with existing buildings, bandwidth, teachers, students, union contracts, school boards…etc.

When we describe blended learning in real schools, or assess outcomes of those implementations, we are reviewing schools, not models. Models are clean, and easily diagrammed. Schools are messy, unpredictable, and differ from one another in important ways.

This is not merely a semantic issue. Anyone who thinks that blended learning is being implemented in existing schools in clean models needs to spend more time in those schools to experience the controlled chaos that describes even the best examples. Using the word “models” when we mean “schools” perpetuates the notion that blended learning implementations are far easier than they really are.