Student Success vs. Patient Survival Rates

Man, the education debate is depressing. Not only is teaching often one of the most difficult jobs one can do - but you have an endless parade of politicians and pundits heaping criticism on your profession after demonstrating again and again that they possess an embarrassing lack of knowledge.

As part of the Obama administration's race to the top, states are being encouraged to use pay-for-performance schemes, a hot concept among current education reformers. Many believe it represents a chance to dramatically alter the nation's troubled schools by providing a sort of market-based, competition-driven prescription for success. If we could only just look at which teachers were doing a good job, and compensate them for it, we would create an atmosphere that rewards success and punishes failure. The process has been likened to other industries where monetary rewards are based on outcomes. Mayor Bloomberg of New York city recently pointed the finger at detractors:

In New York, the state legislature passed a law last year that actually tells principals ‘you can evaluate teachers on any criteria you want, just not on student achievement data.’ That’s like saying to hospitals ‘you can evaluate surgeons on any criteria you want, just not patient survival rates.’

Unfortunately, while the idea is not without merit, it is highly problematic.

Comparing teacher evaluation to patient survival rates is absurd. It displays a fundamental ignorance of basic issues. One is reminded of those who would make generalizations from government spending to household spending. But hey - its just common sense, right?!!

For starters, "patient survival rates" imply highly controlled situations with clear parameters and guidelines to follow. From Wikipedia:

"Relative survival is calculated by dividing the overall survival after diagnosis of a disease by the survival as observed in a similar population that was not diagnosed with that disease. A similar population is composed of individuals with at least age and gender similar to those diagnosed with the disease."

Teacher quality is certainly important - why wouldn't it be? But the problem is in drilling down how to precisely determine that in a way that scales up, allowing districts to set policy. But in any given classroom, the variables effecting student outcomes are immense. Any teacher who has taught at multiple schools with different demographics knows that some classrooms can be orders of magnitude more demanding than others. How then to determine which teacher s performing better?

For example, at school A, the 10th grade Biology class is reading at or above grade level. Most parents can be counted on to enforce academic performance both in and out of class. There are few absences. Students are generally well-fed, groomed and comfortable. Resources are available for field trips (thanks in part to active PTO fund-raising), and most students have access to high-speed internet connections.

At school B however, the 10 grade biology class is reading at the 6th grade level. Few parents can be counted on to enforce performance, many don't return calls or numbers have been disconnected. Absences are frequent, as well as tardies. Students come to class unprepared, nervous and potential for conflict is constant. There are not enough books to send home - and few would be read (or even returned) if they were. There is no money for field trips, requiring the teacher to raise money on their own time. Maybe one or two students have access to a PC at home.

So how would one come up with a formula that would apply to both situations? As any of these variables could rise and fall day to day, how would a results-based compensation system be designed that would be as fair to teachers at school B, as is is at school A?

When you can begin to legitimately respond to these questions, you are ready to enter this debate. Until then, you need to do some more homework.