I've just read an interesting Huffington Post entry from a mathematics teacher in Los Angeles amenable to using student test scores as a factor in his professional appraisal. While his tone is commendably reasonable, both he and all others involved in this debate should beware the assumption of a one-to-one correspondence between teaching inputs and student learning outputs, since other factors might be involved. Two immediate alternative explanations for high student scores immediately come to mind: the (typically Asian) use of tutors or cram schools and the 21st century solution of online tutoring, perhaps through recorded videos like those of the Khan Academy.
Here's a scenario that should give the simple-minded teacher-bonus-for-high-test-score aficianados nightmares, and the rest of us a laugh: two teachers, a retired-on-the-job old-time math teacher who never could teach and a frazzled incompetent rookie science teacher who has received a few days of training before being thrown in front of the toughest students anywhere, are both in hot water because they are about to be eliminated through the new performance appraisal system emphasizing student test scores, for which both are completely unprepared. But luckily into this new situation arrives an immigrant student whose ambitious parent immediately sees that neither of these teachers can teach, hires a tutor, sends her child off to an effective cram school, buys her son extra self-study books, and connects the boy with online tutoring help (perhaps by hacking into the brilliant online tutoring provision of Ontario, Canada); this boy immediately becomes a star student and, being charismatic, is able to turn his friends on to the same resources, and through another parent well capable of manipulating our philanthropic grant system, is able to secure like outside-of-class supports for all of the students in these two incompetent teachers' classes; and so while one teacher sleeps and the other is out of class in psychotherapy, student scores soar, and end-of-year test results determine that these are the most improved classes in the city. An enthusiastic superintendent, dependent upon her dashboards, scans the data and names our two unsuspecting teachers as teachers of the year (they barely beat out a jealous colleague who had helped his students cheat to gain their greatly improved test scores). The superintendent plans a surprise announcement of the awards and invites media cameras in for the presentation of the teacher-of-the-year awards to our two unsuspecting misfits, hoping to see spontaneous displays of highly effective teaching. She enters to find the old-timer asleep and, next door, the science teacher absent, with students studying to compensate for the missing instruction, led by the original star student.
The superintendent is embarrassed and puzzled, and so hires a young university researcher to determine which classroom practices correlate with these higher test scores. The conclusion: lack of teacher interference with student learning, prevented most effectively by absence from school and next best by in-class slumber, correlates most strongly with these students' improvement; and so "Stamp out teaching!" (our faculty was once led to chant "Stamp out literacy" by our school's literacy coordinator in the library of the old Locke High School) becomes the new, evidence-based mantra promoted in all of the schools under our enthusiastic, data-driven superintendent.
This farce might actually work as a movie, where as a comedy it might compensate for the tragedies currently being plotted in our schools.