As I'm sure many of you know, value-added metrics are the latest big thing being touted to evaluate a teacher's worth. Many states and districts have already adopted value-added modeling as a key to teacher evaluation. This clearly brilliant gentleman was explaining to state education leaders, using the familiar gardening analogy, how value-added models worked. He explained how they were getting better all the time and how we would soon have models that were extremely accurate in determining just how much a particular teacher contributed to a particular child's academic growth.

In the gardening analogy, two gardeners are each growing a plant. Gardener A's plant starts a little taller and grows several inches more than Gardener B's plant over the same period of time. But, Gardener A also had more favorable rainfall, soil conditions, and temperatures, so his plant should have grown even more than it did. When those factors were included in the value-added model, Gardener B's plant growth was actually far more impressive, because his plant grew more than would have been predicted given the particular conditions.

Then, I posed a question that completely stumped him. He mumbled and bumbled his way through a non-answer, and then he changed the subject. I was honestly shocked that it seemed he had never considered this extremely important factor in any of his analyses.

Here's the question I asked: Given that these models are being used to measure teacher effectiveness, and they are attempting to control for all the external variables that contribute to a child's academic success or failure, and given that teachers are also expected to implement mandated curricula with a high degree of fidelity and adhering to a strict pacing guide, how are the models controlling for the varying degrees of effectiveness of the curricula and the varying degrees of appropriateness of the curricula for particular students?

The gardening analogy is supposed to comfort teachers who work in schools with more challenges (e.g., poverty, mobility, lack of parent involvement), and it might be appropriate if it actually accounted for all the possible variables. But, we're talking about human beings, not plants, and human responses are not as predictable as plants' responses. And with mandated curricula, we're also talking about a situation equivalent to forcing all gardeners to use the exact same fertilizer and watering schedule, no matter what the conditions call for, and then judging the gardener's value-added based on the results, despite limiting the gardener's capacity to do what gardeners do: use their knowledge and experience to adjust their practices according to what the conditions call for.