We want the identification process to be timely.
We want the identification process to not falsely identify deficient teachers.
This essay will show one method of reconciling that balancing act.
Let us suppose for the moment that the teacher has absolutely no influence on the student's performance. We know this is not true because we would not pay teachers more than minimum wages if any random, custodial adult was just as effective as a trained teacher.
If the teacher has absolutely no influence then in any given cycle the chances of that individual teacher ending up in the lowest performing 10% is exactly 10%. The chances of their ending up in the lowest 10% two cycles in a row is 1%, while their chances of ending up in the lowest 10% three cycles in a row is 0.1%....one chance in a thousand.
But we do know that teachers strongly influence student's ability to learn.
The effect of requiring multiple "failures" of a minimum performance criterion is highly protective of teachers who are on, near, or better than the bubble.
Choice in how many cycles and how to define under-performance is influenced by scarcity of data. Teachers at the elementary level tend to have fewer students (say, a classroom size of 25) and have those students for the entire year. Teachers at the high school level might have 5 classes of 25 and three tri-mesters a year. Clearly the teachers at the high school level generate about fifteen times as many data samples per year as the elementary school teachers and it makes sense to have different under-performance benchmarks for different types of teachers.
How might it work in real life?
Teacher XYZ scores in the lowest 10% (lets call it the under-performance zone) of the teacher population based on the measurement system described in the last post. The administration informs the teacher of the fact.
If Teacher XYZ scores in the under-performance zone in the next cycle, they are place on an improvement plan.
If Teacher XYZ scores in the under-performance zone the third cycle (one chance in a thousand in our null hypothesis) they are released to seek job opportunities that are more in line with their capabilities.
Tuning the system
Natural attrition runs somewhere between 3% and 5% per year.
It seems reasonable that an additional 1%-to-2% per year of non-performance based attrition will not cause undue hiring stresses. If historical data is available, then adjusting the size of the under-performance zone to put 1%-to-2% of the teachers at risk every year is probably good business.
While 1%-to-2% per year seems like a glacial pace to parents angry about school district performance the program throws a very long shadow.
Some teachers will resign when they see the handwriting on the wall.
Most teachers will step up their game to stay out of the 10% non-performance zone. It is like the old joke about out running a grizzly bear.
Finally, eliminating the bottom 10% is HUGE. It is almost impossible for a student to avoid the bottom 10% during their 13 years of education. Clearing out that bottom 10%, even if it takes five or ten years, is a huge step forward.
The non-performance zone
There are some teachers who are so toxic to the system that the school district cannot afford three cycles. There are many reasons why teachers, heck, any employee, can become toxic.
|Approximately 5% of the US population suffers from severe mental illness and highly dependent on medicine to keep the bubble centered. Things are usually OK as long as they take their meds.|
|In round numbers, between 6% and 10% of the population identify as heavy alcohol users. That does not even consider prescription drug abusers and other recreational drug users.|
|Approximately 3% of the population between ages 25 and 45 are undergoing a divorce at any given time.|
Any incursion into the non-performance zone demands attention. Tolerating abysmal non-performance is enabling and is not doing the teacher, or the students any favors.
Perhaps they need counseling. Perhaps they need a 12 step process. Perhaps they need a leave of absence to get their head back together. Regardless, they need help and they need it quickly.