Remember when your teacher would say “Okay someone is coming in today to observe so everyone be on your best behavior” and proceeded to act a little (or a LOT!) differently than usual? How about instead of in-person evaluations schools bought small, cheap cameras to record a few (maybe 3) lessons scattered throughout the year? The mere presence of an “observer” changes the teacher. It would take a lot of guts to show a movie for 3/4 of the class while someone was observing, but on a random school day plenty of teachers will occasionally do so in place of an actual lesson.
I’m glad about the student evaluations (Adults need to trust students! They are not clueless. When they liked a teacher who challenged them, rewarded them, and inspired them– it really is a trustworthy measure), I hope that the test score increases are measured on an accurate scale. Bringing students from 88% proficient to 92% proficient is VERY DIFFERENT than bringing them from 52% proficient to 56% proficient; hopefully these measures reflect that. It could be as simple as the following equivalencies:
1pt. gain 90-100 =
2pt. gain 80-89 =
3pt. gain 70-79 =
4pt. gain 60-69 =
5pt. gain 50-59 =
6+ pt gain 0-49
You would measure the END percentage of proficient students in a class (or the average of their test percentiles, which would be even more specific) and calibrate the system across classes to that.